How to Prepare an Invention Disclosure for Your Patent Attorney

Consider Dr. Elena Marchetti, a battery engineer who has spent two years coaxing a fast-charging method for electric-vehicle cells out of a wall of failed experiments. Her approach is genuinely novel, her test data is strong, and her company is finally ready to spend real money on patent protection. When the patent attorney asks her for an invention disclosure, Marchetti faces a fork in the road that will quietly determine how much that patent is ever worth—though no one will tell her so, and the fork is invisible from where she stands.

She can dash off two paragraphs the night before her intake meeting, the way she'd knock out an expense report. Or she can treat the disclosure as what it actually is: the raw material from which her patent will be forged, the source text of a legal instrument her employer may lean on for twenty years. The difference between those two documents is the difference between a patent competitors design around over a long lunch and one that fences off a market for the life of the grant. The attorney is good. The attorney cannot save a thin disclosure, because an attorney cannot claim what was never disclosed, enable what was never explained, or distinguish prior art the inventor never mentioned.

That is the uncomfortable arithmetic at the center of patent practice, and it is the reason this guide exists. Whether you work in corporate R&D, an academic lab, or a two-person startup, the disclosure you hand counsel sets the ceiling on everything that follows. This guide teaches you how to raise that ceiling: what information patent claims actually require, how to describe a problem and a solution in language that survives translation into claims, why documenting alternatives matters more than almost anyone expects, and how to disclose prior art so it strengthens your application instead of sabotaging it. It also pulls back the curtain on the legal doctrines the disclosure feeds—conception, reduction to practice, inventorship, the duty of candor, enablement, and the timing bars—because once you see how the machine works, you understand what to put in the hopper. For the lay of the broader land, our patent basics and general information concerning patents guide is the natural companion to this one; for what happens after filing, see our guide to responding to patent office actions.

The Disclosure Is the Whole Patent in Miniature

A patent application looks like one document but is really a system of interlocking parts, and every part traces back to the disclosure. Understanding that anatomy is the first step to writing a good one.

The claims are the heart. A claim is a single (often very long) sentence that defines the legal boundary of the patent—the fence line competitors may not cross. Everything else in the application exists to support and justify the claims. The specification, or written description, describes the invention in prose and must satisfy two distinct statutory requirements under 35 U.S.C. § 112: it must enable a person of ordinary skill in the art to make and use the invention without undue experimentation, and it must demonstrate that the inventor actually possessed the full claimed invention. Drawings carry the visual load that words handle badly—components, arrangements, signal flows. The abstract is a short summary for searching and classification. And underneath all of it sit the inventor's own records: the lab notebooks, the dates, the data.

Each of these flows from the disclosure, and none can rise above it. Marchetti's attorney cannot claim a charging method the disclosure never describes, cannot enable an embodiment the disclosure never explains, cannot illustrate a circuit the disclosure never sketches, and cannot prove a conception date the disclosure never anchors. The patent process is a manufacturing process: better raw material, better finished product. There is no alchemy step where thin input becomes a thick patent.

When disclosures are vague, thin, or disorganized, the damage shows up in predictable places. Without a grasp of the invention's full scope—the alternatives, the variations, the boundary cases—counsel drafts claims narrower than necessary, handing competitors room to engineer around them. Gaps in the disclosure become gaps in the specification, and a specification that fails to support a claim element is an invitation to invalidation for inadequate written description or lack of enablement under § 112. Complex projects frequently hide several patentable inventions, and an incomplete disclosure may capture only the obvious one while a subsidiary innovation worth protecting in its own right slips away unnamed. There is a direct financial cost, too: patent attorneys bill by the hour, and every follow-up call to extract information that should have been in the disclosure is time the client pays for at a partner's rate. Finally—and this is the one that can detonate years later—the duty of candor obliges everyone substantively involved in prosecution to disclose material prior art they know about. A disclosure that buries a known reference risks later discovery of the omission and, with it, a holding of inequitable conduct that renders the entire patent unenforceable. We will return to that bomb; for now, hold the thought that honesty in the disclosure is not just good manners.

Practical Law's own guidance to patent counsel makes the stakes explicit in a way inventors rarely hear: the confidential invention disclosure "is often the first full record of the invention, which may be significant in patent prosecution and possible litigation over many years," and "access to invention contributors and their records can change." People leave companies. Memories fade. The notebook gets boxed up. The disclosure is the time capsule, and you are writing it for a litigator you have never met who will open it in 2034.

What Claims Actually Require From You

Because the claims are the point, the fastest way to understand what to put in a disclosure is to understand what claims need. A claim defines an invention in terms of elements—components or steps—and the relationships among them. Independent claims stand on their own. Dependent claims incorporate every limitation of the claim they refer back to and then add at least one more, narrowing scope while gaining a measure of safety. A patent is typically a nested family of claims, broad at the top and progressively narrower below, so that if a court or examiner knocks out the broadest claim, narrower ones survive.

A representative independent claim in Marchetti's field might read like this:

A system for charging a battery comprising: a power stage configured to deliver current to the battery in pulses; a temperature sensor configured to measure a temperature associated with the battery; a thermal model configured to predict an internal cell temperature from a pulse parameter and the measured temperature; and a controller configured to adjust the pulse parameter based on the predicted internal cell temperature.

Four elements—power stage, temperature sensor, thermal model, controller—each defined by what it does and how it relates to the rest. A dependent claim then tightens the screws:

The system of claim 1, wherein the controller adjusts a pulse duration and a rest period to hold the predicted internal cell temperature below a degradation threshold.

To draft claims at the right scope, counsel needs several categories of information, and an inventor who supplies them turns a frustrating intake interview into a fast one. The table below is a working checklist—keep it beside you while you draft, rather than reading it once and moving on.

Claim scope is a perpetual tradeoff, and it helps to feel the tension before you draft. Broader claims reach more infringers but expose more surface to prior-art attack; narrower claims survive more easily but are simpler to design around. The best disclosures let counsel claim at several altitudes at once. At the broadest altitude is the most general description that still captures the invention's essential character—the minimum set of elements that produces the technical benefit. At an intermediate altitude sit the preferred features that make the invention valuable without committing to every screw. At the narrowest altitude is the specific, fully featured embodiment Marchetti actually built and tested, which becomes the fallback if the broad claims are rejected during prosecution. Practical Law puts the same point to patent counsel as an instruction: gather "information sufficient to prepare a specification and claims set, with the prior art in mind, that describes the invention in both broad and progressively narrower terms." Your job as the inventor is to give counsel the material to do that. You do not draft the claims. You make good claims possible.

Describe the Problem, Then the Mechanism—Not Just the Result

Patents are, at bottom, solutions to problems, and a disclosure should open by naming the problem before explaining the cure. Skipping the problem is the single most common rookie move, and it is costly in a specific, doctrinal way.

A crisp problem statement does real legal work on the non-obviousness question under 35 U.S.C. § 103. One of the most persuasive ways to show an invention was not obvious is to establish that the problem was recognized in the field and stubbornly unsolved—that smart, motivated people stared at it and failed. Courts call this long-felt but unmet need, and it is among the secondary considerations of non-obviousness that the Federal Circuit takes seriously. (The framework traces to Graham v. John Deere Co., 383 U.S. 1 (1966), which still governs the obviousness inquiry, and the Supreme Court's later loosening of the "motivation to combine" rule in KSR International Co. v. Teleflex Inc., 550 U.S. 398 (2007), only makes objective evidence of need more valuable to an applicant.) A vivid problem statement also guides claim scope—it points counsel at the essence of the solution rather than at incidental trim—and it becomes the "background of the invention" that opens the specification. The difference between a weak and a strong statement is concreteness:

Once the problem is on the page, explain how the invention solves it—and here the cardinal rule is to describe the mechanism, not just the result. It is not enough to say the system is faster. The disclosure must say why it is faster: what architectural change, algorithmic step, or structural choice produces the gain. The most useful disclosures connect each technical choice to its benefit in an explicit chain of cause and effect. Marchetti should write that by applying current in high-amplitude pulses with rest intervals between them (the choice), her method lets accumulated heat dissipate during the rests (the mechanism), which permits a higher average current without crossing the degradation threshold (the result). Carry enough technical detail that someone skilled in the field could actually build it: if the invention turns on an algorithm, explain how the algorithm runs; if on a structure, describe the parts and their arrangement; if on a process, walk the steps in order.

A simple discipline keeps you honest. For each significant aspect of the invention, answer both how it works and why it works better than the alternatives. Applied to Marchetti's predictive thermal control, the "how" is that a model estimates internal cell temperature from pulse parameters and the measured surface temperature, then adjusts pulse duration and rest period in real time; the "why" is that prior systems reacted to surface temperature only after heat had already built up, forcing conservative current limits, whereas predicting internal temperature lets the controller ride right up against the safe boundary and therefore charge faster at the same level of cell safety. Notice that the "why" doubles as a built-in distinction over the prior art—which is exactly the kind of material counsel mines when an examiner pushes back.

One caution borrowed from seasoned patent counsel: while the disclosure should be candid and complete about how the invention works, it is wise to discuss any limits, drawbacks, or unsuccessful avenues orally with your attorney rather than memorializing them in colorful prose. A throwaway written line like "this approach completely fails above 60°C" can be quoted back at you in litigation years later, stripped of context. Tell your lawyer everything; let your lawyer decide what belongs in the written record.

Document the Alternatives You Didn't Build

If there is one habit that separates a patent with teeth from a patent that decorates a wall, it is the disclosure of alternative embodiments. The most common failure in invention disclosures is to describe only the one specific thing the inventor built—the prototype on the bench—with no account of the other ways it could have been built. Narrow focus breeds narrow patents, because claims drawn from a single described implementation are vulnerable to being construed as limited to that implementation. Describe alternatives, and counsel can claim the concept rather than one execution of it; doing so closes off design-arounds by sweeping competing approaches inside the fence; and it stocks the cupboard with fallback positions, because alternatives disclosed in the specification can be pulled into amended claims when a broad claim collides with prior art during prosecution.

The crucial—and counterintuitive—point is that you do not need to have built the alternatives. U.S. patent law imposes no requirement that every disclosed or claimed embodiment be actually reduced to practice. A "constructive" reduction to practice—a description detailed enough to enable a skilled artisan—suffices for the specification. The Supreme Court confirmed as much in Pfaff v. Wells Electronics, Inc., 525 U.S. 55 (1998), holding that an invention can be "ready for patenting" by either actual reduction to practice or by drawings and descriptions sufficiently specific to enable a person skilled in the art to practice it. Practical Law's invention-disclosure guidance drives the point home with an explicit instruction to contributors: "Do not limit the description to the embodiments or experiments that have already actually been conducted." So when you brainstorm alternatives, you are not lying about what you built—you are mapping the design space you understand.

There are several flavors of alternative worth capturing. Component alternatives ask what other parts could perform the same function—Marchetti's controller might be a dedicated microcontroller, an FPGA, or a function folded into the vehicle's existing battery-management processor. Process alternatives ask what other steps or sequences reach the same end—her thermal model could run continuously, on a fixed schedule, or only when measured temperature crosses a trigger. Configuration alternatives ask what other arrangements of the same parts are possible—running the prediction locally on the charger versus offloading it to a connected vehicle. Parameter alternatives ask what ranges of values could be used—pulse durations from 100 to 500 milliseconds, rest periods from 50 to 200 milliseconds, peak currents up to 4C—each with a stated preferred value and an explanation of the tradeoffs across the range.

Patent attorneys formalize this instinct as the relationship between a genus (a broad category) and its species (the specific examples within it). For each aspect of an invention, name the broadest category that captures your approach, then list specific examples within it and flag the one you prefer. The pattern travels across every field:

There is, however, a doctrinal subtlety here that the best inventors internalize, because it tells you how to write the alternatives, not just that you should. A broad genus claim is only as good as the specification's support for it. Under the written-description and enablement requirements of § 112, a single disclosed species supports a broad genus claim only if the disclosure gives a skilled artisan enough guidance to identify the other members of the genus without undue experimentation. The Federal Circuit allowed exactly that kind of single-species-supports-genus claim where the guidance was adequate in Hynix Semiconductor Inc. v. Rambus Inc., 645 F.3d 1336, 1352 (Fed. Cir. 2011), and refused it where a single species had unique, unpredictable properties in University of Rochester v. G.D. Searle & Co., 358 F.3d 916, 926 (Fed. Cir. 2004). The Supreme Court drew the same line for enablement in Amgen Inc. v. Sanofi, 598 U.S. 594 (2023): an antibody patent that defined its genus functionally—by what the antibodies do—but disclosed only twenty-six specific antibodies and a "roadmap" for finding the rest did not enable the full claimed scope, because the roadmap amounted, in the Court's memorable phrase, to "little more than two research assignments." The lesson for your disclosure is concrete: when you list a genus, give counsel the bridges between the species—the design rules, the selection criteria, the worked examples—so that the claim you eventually obtain is supported and not merely aspirational. A disclosure that says "the model could be any machine-learning architecture" is weaker than one that explains which architectures work, why, and how a skilled person would choose among them.

This scope-mapping exercise also feeds directly into a freedom-to-operate analysis, because the alternatives you identify are often exactly the design space your competitors occupy—and the same map that broadens your claims tells you where you might be at risk of infringing someone else's.

Identifying and Disclosing Prior Art (and Why Hiding It Can Detonate Your Patent)

Prior art—the existing body of knowledge relevant to whether an invention is patentable—drives most of what happens during prosecution, and inventors usually know the prior art in their field better than anyone else alive. Capturing that knowledge in the disclosure is both strategically smart and, in part, legally mandatory.

The legal mandate is the duty of candor, codified at 37 C.F.R. § 1.56, which obliges everyone substantively involved in prosecuting an application—each named inventor, each attorney or agent, and anyone else with a hand in it—to disclose information material to patentability. Each inventor personally acknowledges this duty by signing the inventor's oath or declaration (37 C.F.R. § 1.63(c)). The consequence of breaching it is the harshest sanction in patent law. A holding of inequitable conduct renders the entire patent unenforceable—not just the tainted claim, the whole patent—and the taint can spread to related patents in the same family. The Federal Circuit, sitting en banc in Therasense, Inc. v. Becton, Dickinson & Co., 649 F.3d 1276 (Fed. Cir. 2011), tightened the standard precisely because courts had been finding inequitable conduct too freely. A challenger must now prove, by clear and convincing evidence, both but-for materiality (the patent would not have issued had the PTO known the truth) and a specific intent to deceive—and the intent to deceive must be "the single most reasonable inference" from the evidence, not merely a reasonable one. There is also a narrow exception for affirmative egregious misconduct—filing an unmistakably false affidavit, say—which is material regardless of the but-for test (Therasense, 649 F.3d at 1292–93; see Intellect Wireless, Inc. v. HTC Corp., 732 F.3d 1339 (Fed. Cir. 2013), invalidating a patent over a false declaration about an invention's reduction to practice).

Two practical points follow. First, Therasense makes inequitable conduct harder to prove, but it does not make the duty optional—the rational strategy remains to disclose known material prior art rather than gamble that an examiner overlooks it, because the downside of a successful charge is total and the upside of concealment is nil. Second, your job in the disclosure is not to decide what is "material"—Practical Law's disclosure form expressly tells contributors, "Do not make any conclusions concerning patentability." You list what you know; counsel decides what to cite to the PTO in an Information Disclosure Statement. (Our deep dives into inequitable conduct in patent prosecution and finding evidence of inequitable conduct trace the doctrine in full.)

Prior art is also broader than most inventors assume. It includes issued patents and published applications from the United States and foreign offices; academic publications—journal articles, conference papers, theses, dissertations; commercial products on sale that embody the technology; published technical standards; public demonstrations, including trade-show booths and posted videos; and—this is the one that catches people—the inventor's own earlier work. Your own publication, your own conference talk, your own product launch, your employer's prior sale: any of these can become prior art against your own application if it predates your filing by more than the grace period. We will unpack the timing bars below; for the disclosure, the rule is to list your own public activities as carefully as anyone else's.

A disclosure should do three things with prior art. It should list the specific references you know of, with enough identifying detail—patent numbers, author names, titles, links—for counsel to find them. For each, it should explain how the invention differs, which lets counsel anticipate rejections and prepare distinguishing arguments. And it should flag where prior art probably exists but you haven't looked, so counsel can decide whether to commission a professional patentability search. A compact table keeps it organized:

Certain categories of prior art deserve a hard second look because inventors routinely forget them: art cited in related applications such as continuations, parents, or foreign counterparts; references you reviewed while developing the invention; prior art mentioned in passing by colleagues, supervisors, or collaborators; and your own prior disclosures. A useful self-test is to ask, "What did I read, see, or buy that pointed me toward this idea?"—and then write all of it down.

The Hidden Spine: Conception, Reduction to Practice, and the Lab Notebook

Behind every good disclosure stands a good set of records, and the law that governs those records is worth understanding because it explains why the dates and the notebook entries matter. U.S. invention is a two-act play: conception followed by reduction to practice.

Conception is the legal heart of invention—"the formation in the mind of the inventor of a definite and permanent idea of the complete and operative invention" (Invitrogen Corp. v. Clontech Laboratories, Inc., 429 F.3d 1052, 1063 (Fed. Cir. 2005)). It is the "aha." Critically, conception requires a complete and operative idea: a notebook entry proving conception must encompass every feature of the eventual claim and must let an unconnected skilled artisan reproduce the invention without extensive experimentation (Singh v. Brake, 317 F.3d 1334, 1340 (Fed. Cir. 2003)). Conception does not require the inventor to know in advance that the invention will work—the Federal Circuit held in Dana-Farber Cancer Institute, Inc. v. Ono Pharmaceutical Co., 964 F.3d 1365, 1372 (Fed. Cir. 2020), that an inventor need not have proven the invention's success to have conceived it.

Reduction to practice is the second act: either actual reduction to practice—building a working embodiment and confirming it is suitable for its intended purpose (Medichem, S.A. v. Rolabo, S.L., 437 F.3d 1157, 1169 (Fed. Cir. 2006))—or constructive reduction to practice, which is simply filing a patent application that enables the invention. Those who merely build or test under an inventor's direction are not, by that act alone, inventors; conception is the touchstone of inventorship, not construction.

You might reasonably ask why any of this still matters under the America Invents Act's first-inventor-to-file system, where the filing date usually governs. It matters for several reasons that bear directly on the disclosure. Invention dates still surface in derivation proceedings (did someone steal your idea and file first?), in disqualifying prior art under the grace-period exceptions, and in inventorship disputes. And the records that establish those dates—the lab notebooks—are the corroboration without which an inventor's later testimony is, as the courts bluntly put it, unreliable. Uncorroborated inventor testimony about conception or prior invention "is unreliable" (James v. J2 Cloud Services, LLC, 823 F. App'x 945, 949 (Fed. Cir. 2020)), and an unsigned, unwitnessed notebook "is given no more weight than the inventor's testimony" itself (Reese v. Hurst, 661 F.2d 1222, 1231 (C.C.P.A. 1981)). Conversely, a notebook signed and witnessed by a non-inventor skilled in the art who watched the work being done can be powerful corroboration, sometimes admissible without more (Medichem, 437 F.3d at 1169–70). The Federal Circuit invalidated an inventorship claim for want of corroboration in Apator Miitors ApS v. Kamstrup A/S, 887 F.3d 1293 (Fed. Cir. 2018), precisely because the inventor's evidence rested on his own unwitnessed records.

The practical translation for your disclosure—and for the daily discipline that feeds it—is this. Record your work as you go, sign and date each entry, and have a qualified, non-inventor witness sign too, near the time of the work. Record the failures alongside the successes, in the same level of detail; an unbroken record is more credible than a curated one. Cross-reference related notebooks and explain gaps in the dates. Keep legal characterizations out of the notebook—do not editorialize about patentability, do not disparage competitors, and do not transcribe your conversations with counsel, because the notebook may someday be evidence and you do not want to hand the other side a quotable misstep. (The Federal Circuit has found inequitable conduct where misleading characterizations crept into the technical record; see Hoffmann-La Roche, Inc. v. Promega Corp., 323 F.3d 1354 (Fed. Cir. 2003).) When you sit down to write the disclosure, the notebook is your source document: it supplies the conception date, the development milestones, and the data that will eventually support your claims. Build a trade secret protection program and good notebook hygiene early, and the disclosure practically writes itself.

Inventorship: A Legal Determination, Not a Courtesy

Few patent mistakes are as quietly fatal—or as socially awkward—as getting inventorship wrong, and the disclosure is where the facts that fix it are first captured. Naming the wrong people can invalidate a patent or, at minimum, cloud its ownership, because in the United States ownership flows from inventorship.

The legal standard begins with conception, just discussed. An inventor is someone who contributed to the conception of at least one claim. That last phrase is doing a lot of work. The Federal Circuit's canonical test comes from Pannu v. Iolab Corp., 155 F.3d 1344, 1351 (Fed. Cir. 1998): a putative joint inventor must (1) contribute in some significant manner to the conception or reduction to practice of the invention, (2) make a contribution to the claimed invention that is not insignificant in quality when measured against the full invention, and (3) do more than merely explain to the real inventors well-known concepts or the current state of the art. A joint inventor need not contribute to every claim, need not work at the same time as the others, and need not make a contribution of equal magnitude—but the contribution must clear the Pannu bar, and it must go to the conception, not the mere carrying-out, of the invention. Each joint inventor, the statute says, applies jointly (35 U.S.C. §§ 116, 256, the latter providing the mechanism to correct inventorship later).

Several roles that feel like inventorship are not. Suggesting a problem to solve, supplying funding or lab resources, running experiments under another's direction, providing well-known background, or supervising without making a conception-level technical contribution do not, without more, make a person an inventor. Marchetti's department head who approved the budget and her technician who ran the charge-cycle tests to her specification are valued colleagues and may be the reason the project succeeded—but neither is an inventor unless he or she contributed to conceiving a claimed feature. The temptation to add a generous name (the boss, the funder, a beloved postdoc) or to omit an inconvenient one (a departed collaborator, an outside consultant) is strong and dangerous. A named inventor is presumed to be a true one (Cumberland Pharmaceuticals Inc. v. Mylan Institutional LLC, 846 F.3d 1213, 1218 (Fed. Cir. 2017)), and an inventorship error made innocently can be corrected under § 256, but an error made deceptively can render the patent unenforceable (Perceptive Biosystems, Inc. v. Pharmacia Biotech, Inc., 225 F.3d 1315, 1322 (Fed. Cir. 2000)).

So the disclosure should not pronounce on inventorship—that is counsel's call once the claims take shape—but it should give counsel the raw facts to make the call. Document what each contributor did, in detail, cross-referenced to the dates and records. Practical Law's disclosure form deliberately gathers "contributors" (anyone who created, discovered, developed, tested, or analyzed any aspect of the invention) precisely so counsel can sort inventors from non-inventors later, and so that non-inventor contributors who hold useful data and documents can be found. A single disclosure covering the whole invention usually beats several partial ones; one inventor should be designated to coordinate and serve as counsel's primary contact; and all potential inventors should review the disclosure for accuracy before submission. Where inventors sit at different organizations—a university–industry collaboration, say—the disclosure must be handled consistently with each institution's IP policies, and ownership will turn on the assignment provisions of each contributor's agreements. (This is exactly why robust employee invention assignment agreements matter so much: inventorship sets who invented, but the assignment agreement sets who owns, and a gap between the two is how companies lose patents they paid to develop.)

A note on the frontier: as AI tools migrate from the office into the lab, the question of whether a machine's contribution counts toward conception—and what to do when no human conceived a claimed feature—is no longer hypothetical. Under current U.S. law an inventor must be a natural person, but the line between human-directed and machine-generated conception is getting blurry. We take that up at length in AI-generated inventions: who owns what the machine creates and in our survey of artificial intelligence and inventorship. If an AI system played a meaningful role in conceiving any feature, say so in the disclosure—do not paper over it—and let counsel navigate the inventorship and ownership consequences.

Timing, Confidentiality, and the Bars That Can Erase Your Rights

Timing can be decisive in patent law because certain events slam the door on patentability, and the disclosure is where the dates that govern those events are first pinned down. The United States runs on a first-inventor-to-file system under the America Invents Act, which applies to any application with a claim having an effective filing date on or after March 16, 2013. As between competing inventors, the one who files first generally wins—a sharp break from the old first-to-invent regime, which is why the AIA also created derivation proceedings to police outright idea-theft.

The events that bar a patent come from 35 U.S.C. § 102. Under the AIA, prior art includes anything "patented, described in a printed publication, or in public use, on sale, or otherwise available to the public" before the effective filing date—anywhere in the world. U.S. law softens this with a one-year grace period for certain of the inventor's own disclosures: publish the invention, present it at a conference, post it online, demonstrate it publicly, or offer it for sale, and you start a one-year clock, after which the right to a U.S. patent is lost. Two features of this regime catch inventors off guard, and both belong in the disclosure.

First, the on-sale and public-use bars are broader and stranger than intuition suggests. The Supreme Court held in Helsinn Healthcare S.A. v. Teva Pharmaceuticals USA, Inc., 586 U.S. 123, 139 S. Ct. 628 (2019), that even a secret, confidential commercial sale—a deal whose terms are under NDA and that discloses nothing about how the invention works—can trigger the on-sale bar. The AIA's addition of the words "or otherwise available to the public," the Court ruled, did not change the settled meaning of "on sale"; a commercial sale need not make the invention public to bar it. The on-sale bar attaches when the invention is the subject of a commercial offer for sale and is "ready for patenting" under the Pfaff test discussed earlier (Pfaff v. Wells Electronics, Inc., 525 U.S. 55, 67 (1998))—and a single offer can do it (Special Devices, Inc. v. OEA, Inc., 270 F.3d 1353 (Fed. Cir. 2001)). The public-use bar is similarly capacious: even an inventor's secret commercial use of a process can count, an old rule from Judge Learned Hand's Metallizing Engineering Co. v. Kenyon Bearing & Auto Parts Co., 153 F.2d 516 (2d Cir. 1946), that the Federal Circuit reaffirmed in BASF Corp. v. SNF Holding Co., 955 F.3d 958 (Fed. Cir. 2020). There is a genuine experimental-use exception—bona fide, non-commercial testing to perfect the invention does not count—but it is narrow and does not cover testing for marketing, regulatory, or customer-acceptance purposes (Allen Engineering Corp. v. Bartell Industries, Inc., 299 F.3d 1336, 1353 (Fed. Cir. 2002)). The takeaway for the disclosure: list every offer for sale, every demonstration, every beta deployment, and every quote you've sent a customer, with dates—because what looks like a routine business deal to you may look like a statutory bar to a litigator.

Second, the grace period is largely an American indulgence. Most of the world demands absolute novelty: any public disclosure anywhere before filing destroys patent rights there. An inventor who may want protection in Europe, Japan, China, or elsewhere should therefore file before any public disclosure, rather than leaning on a grace period most jurisdictions do not honor. This single fact reshapes the timing strategy for anything with international ambitions: file first, publish second.

These rules make confidentiality and promptness not just prudent but essential. Before filing, keep the invention confidential—mark the disclosure document itself confidential (it is, after all, an attorney-client privileged communication when prepared for counsel), and limit who sees it. Where research will be published, coordinate with counsel to file before publication; many organizations mandate patent review before any public talk or paper, precisely to avoid an inadvertent bar. Disclosures made under a properly drafted non-disclosure agreement generally do not trigger the public-disclosure bar—an NDA's confidentiality obligation is one of the factors courts weigh in deciding whether a use was "public"—but reliance on an NDA demands careful documentation, and Helsinn is a reminder that a confidential sale is a different animal from a confidential disclosure. Two simple rules of thumb: submit the disclosure as soon as the invention is developed enough to describe, rather than waiting for an imminent product launch; and if a potentially barring event has already happened, calendar the one-year deadline immediately and make sure filing beats it. For confidential technical information you may ultimately decide to keep as a trade secret rather than patent—a real fork in the road, since patenting requires public disclosure while trade-secret protection requires the opposite—see trade secrets in the age of remote work and cloud computing.

Best Mode and the Subtle Demands of § 112

One more statutory requirement shapes what belongs in a disclosure, and it confuses inventors and lawyers alike: the best-mode requirement. Section 112(a) still requires that the specification "set forth the best mode contemplated by the inventor of carrying out his invention." If, at the time of filing, you believe one particular implementation is best—a specific catalyst, a specific algorithm, a specific tolerance—you must disclose it; you may not keep your secret sauce out of the patent while claiming the territory. So the disclosure should tell counsel which embodiment you consider preferred and why, holding nothing back.

Here is the wrinkle that trips people up. The AIA (effective for proceedings filed on or after September 16, 2011) eliminated failure to disclose the best mode as a basis for invalidating or holding a patent unenforceable (35 U.S.C. § 282(b)(3)(A)). Best mode remains a requirement of patentability that the examiner can in theory police, but a defendant in litigation can no longer attack a patent for a best-mode lapse. This does not license you to withhold your preferred embodiment—an intentional concealment can still surface as evidence in an inequitable-conduct narrative, and a specification that hides the best implementation may fail the related enablement and written-description tests anyway. The clean approach is the honest one: disclose your best mode in the disclosure, and let counsel ensure the specification carries it.

Which brings us back to the twin pillars of § 112, because the disclosure is where their fate is decided. Enablement asks whether the specification teaches a skilled artisan to make and use the full scope of the claims without undue experimentation—the eight-factor inquiry of In re Wands, 858 F.2d 731 (Fed. Cir. 1988), and the full-scope rigor of Amgen v. Sanofi. Written description asks whether the specification shows the inventor possessed the full claimed invention, the "possession" test of Ariad Pharmaceuticals, Inc. v. Eli Lilly & Co., 598 F.3d 1336 (Fed. Cir. 2010) (en banc). Both are decided on the contents of the application, and the application is built from your disclosure. If you claim a range of pulse durations from 100 to 500 milliseconds, the disclosure had better describe and enable that whole range, not just the 200-millisecond pulse you happened to test—because a claimed range unsupported by the disclosure is exactly the kind of vulnerability the Federal Circuit struck down in Eiselstein v. Frank, 52 F.3d 1035 (Fed. Cir. 1995). The discipline of describing alternatives, ranges, and worked examples is not busywork; it is what makes broad claims survivable.

Tailoring the Disclosure to the Type of Invention

Different inventions stress different parts of the disclosure, and a little tailoring goes a long way.

Software and algorithm inventions live under the shadow of Alice Corp. v. CLS Bank International, 573 U.S. 208 (2014), which held that abstract ideas implemented on generic computers are not patent-eligible under 35 U.S.C. § 101. The antidote, as the Federal Circuit's eligibility cases show, is to frame the invention as a technical solution to a technical problem rather than a business solution to a business problem. Emphasize concrete improvements to the functioning of the computer itself—processing efficiency, memory use, security—rather than improved business outcomes; the patents that survived Alice did so because they were, in the court's words, "necessarily rooted in computer technology" (DDR Holdings, LLC v. Hotels.com, L.P., 773 F.3d 1245, 1257 (Fed. Cir. 2014)) or claimed concrete, rule-based improvements (McRO, Inc. v. Bandai Namco Games America Inc., 837 F.3d 1299 (Fed. Cir. 2016)). For software, the disclosure should describe the system from the computer's perspective—architecture, modules, data structures, the algorithm broken into steps, ideally with flowcharts or pseudocode—and should flag any open-source components, which can constrain how the eventual patent is asserted. Our guide to patent eligibility after Alice develops these strategies in depth.

Mechanical and hardware inventions reward multiple drawing views—perspective, exploded, cross-section—along with preferred materials, critical dimensions, tolerances, and any novel manufacturing method. A disclosure that says "a fastener" is weaker than one that says "a fastener (preferably a captive M3 hex-socket screw; alternatively a quarter-turn cam lock where tool-free service is required)."

Chemical and pharmaceutical inventions require precise molecular structures for claimed compounds, synthesis routes, and—critically—experimental data demonstrating the claimed properties, because in these "unpredictable arts" the written-description and enablement bars sit higher. A disclosure here should also state which structural modifications preserve activity, since that is what supports a claim to a class of compounds rather than a single molecule (a lesson written in blood by University of Rochester and Amgen).

Business-method inventions face the steepest Alice climb. The disclosure should identify the specific technical systems that implement the method, emphasize the technical processes the method improves, and explain why the invention is more than an abstract idea running on generic hardware—because absent that technical anchor, the claim is at grave risk under § 101.

Working Within a Corporate Disclosure Program

In many companies the disclosure does not go straight to outside counsel; it runs a gauntlet first. A patent committee weighs it against criteria like these: the technical merit and soundness of the advance; its alignment with business strategy and whether it protects products the company sells or plans to sell; the surrounding patent landscape and whether there is freedom to operate; the practicality of enforcement (could infringement even be detected, or would a competitor's use be invisible inside a black box?); and a cost-benefit assessment of whether the expected protection justifies the cost of prosecution and decades of maintenance fees. Some organizations formalize this into a scoring matrix:

An inventor inside such a program is well served by treating its rituals as allies rather than red tape. Meet the disclosure deadlines—they often exist precisely to ensure filing before a statutory bar. Complete the standard forms thoroughly rather than minimally; those forms were designed to capture exactly what the committee and counsel need, and a half-filled form just generates follow-up. Supply business context that helps the committee prioritize your disclosure against the dozen others competing for the same budget. Be ready to present and defend the disclosure—to explain the mechanism, the alternatives, and the prior art—if the committee asks. And follow up on status, providing additional information promptly. The committee is not the enemy of your patent; it is the body that decides whether your patent gets funded.

From Disclosure to Issued Patent

It helps to see the whole arc, because every step downstream draws on the disclosure. On receiving it, the patent attorney reviews it to understand the invention and its potential claim scope, conducts or commissions a prior-art search that goes beyond what you disclosed, develops a claim strategy, and drafts the claims, specification, abstract, and drawings. The attorney then sends drafts back to the inventors for a technical-accuracy review—a step inventors should take with deadly seriousness, because attorneys, however skilled, occasionally garble a technical detail, and catching the error before filing is trivial while fixing it after filing can be impossible. Then the application goes to the USPTO. (Before any of this, counsel may run a patentability search and opinion to gauge the odds and to satisfy the prudent practice of knowing the prior-art landscape going in.)

After filing, an examiner takes up the application, and the first office action commonly rejects at least some claims—over prior art, under § 101, or for § 112 defects. This is utterly routine; most applications draw an initial rejection, and a rejection is the start of a negotiation, not a verdict. Counsel responds by arguing against improper rejections and, where appropriate, amending claims to overcome the legitimate ones. Additional rounds may follow until the examiner allows the claims or issues a final rejection. The whole dance typically runs two to four years, and a strong disclosure pays dividends the entire way, because the alternatives, distinctions, and worked examples it captured are exactly what counsel reaches for to overcome an obviousness or anticipation rejection. The mechanics of that back-and-forth are the subject of our guides to responding to patent office actions and overcoming obviousness rejections. And once a patent issues, the careful scoping you did at the disclosure stage informs everything downstream, including decisions about licensing the patent.

A Reusable Invention Disclosure Form

The template below captures what counsel needs to draft a strong application. It is a reference tool—fill it in, adapt it to your organization's requirements and your invention's complexity, and treat the prompts as a checklist rather than prose to be read straight through. Complete it in consultation with counsel where possible, mark it confidential, and remember the form's own first instruction: do not draw conclusions about patentability or inventorship, and do not disclose the invention publicly without counsel's approval.

INVENTION DISCLOSURE FORM

CONFIDENTIAL — ATTORNEY-CLIENT PRIVILEGED COMMUNICATION

Section 1: Administrative Information

Section 2: Invention Summary

A brief (1–2 paragraph) plain-language summary for a non-technical reader. What does the invention do, and why is it valuable?

Section 3: Technical Problem

The technical problem the invention addresses. What limitations exist in current approaches, why are they significant, and what is the root cause? Quantify where possible.

Section 4: Technical Solution (How and Why)

How the invention solves the problem—the mechanisms, structures, or processes—in enough detail that a skilled person could implement it. For each significant aspect, explain both how it works and why it works better than the alternatives.

Section 5: Components and Relationships

Section 6: Alternative Embodiments (Do Not Limit to What You Built)

Section 7: Technical Advantages

Section 8: Prior Art You Are Aware Of

Also note: areas where prior art likely exists but you have not searched; references you reviewed while developing the invention; and your own prior publications, presentations, or products.

Section 9: Conception and Development (with Records)

Section 10: Public Disclosures, Sales, and Offers

Section 11: Contributor Roles (for Inventorship)

Counsel will determine inventorship. Identify everyone who created, discovered, developed, tested, or analyzed any aspect of the invention, inside or outside the organization.

Section 12: Drawings, Figures, and Best Mode

Attach labeled drawings, flowcharts, schematics, or photographs. Identify the embodiment you currently consider best, and explain why.

Inventor Certification

I certify that the information in this disclosure is accurate and complete to the best of my knowledge, and I understand my duty of candor to the USPTO.

Inventor signature: _________________ Date: _____________

Effective vs. Ineffective Disclosures: Four Worked Contrasts

The principles above are easiest to absorb by watching the same point done badly and then done well. These contrasts are reference material; they reward a reread while you draft a real disclosure.

Describing the technical problem. Ineffective: "Current battery charging is slow and inefficient." Too vague to be useful—it neither quantifies the problem nor names its cause. Effective: "Lithium-ion EV batteries currently require 30–60 minutes for an 80% charge on DC fast charging. This is a major barrier to adoption, because consumers accustomed to five-minute gasoline refueling perceive EVs as inconvenient for long trips. The primary limitation is heat generation during fast charging—pushing current above roughly 2C causes temperature increases that degrade battery chemistry and trigger thermal-management systems that throttle charging to prevent damage." It quantifies the problem, explains why it matters, names the technical root cause, and gives a concrete parameter that counsel can build a claim around.

Describing the solution. Ineffective: "Our invention charges batteries faster by using a better algorithm." This tells counsel nothing claimable. Effective: "Our invention uses pulsed charging with variable duty cycles optimized by real-time thermal modeling. Rather than applying constant current until thermal limits are reached, we apply high-current pulses (up to 4C) for short durations (100–500 ms) followed by rest periods (50–200 ms) that let heat dissipate. A thermal model running on the battery-management system predicts internal temperature from the pulse parameters and the measured external temperature, dynamically adjusting pulse duration and rest period to hold internal temperature below degradation thresholds while maximizing average current. The result is an effective charging rate equivalent to 3C continuous current while keeping internal temperature within safe limits." It explains the mechanism, gives specific parameters and ranges, says why it works, and quantifies the result.

Documenting alternatives. Ineffective: "The sensor could be other types of sensors." Effective: "The position sensor in our preferred embodiment is a quadrature optical encoder providing 1024 counts per revolution. Alternatives include Hall-effect sensors with a magnetic encoding disk (lower resolution but more robust to contamination), potentiometric sensors (lowest cost but limited range and subject to wear), a resolver with a resolver-to-digital converter (highest reliability for safety-critical use), and an inertial measurement unit with rotational integration (suitable where absolute position is not required). The optical encoder is preferred where high resolution at moderate cost is needed, but the resolver is preferable for automotive or aerospace applications requiring safety certification." The effective version names specific species, explains their tradeoffs, and identifies when each is preferred—exactly the guidance § 112 demands to support a genus claim.

Identifying prior art. Ineffective: "I think there might be some patents in this area." Effective: a short, referenced rundown—Smith's 2020 patent describes pulsed charging with fixed parameters, distinguished by Marchetti's dynamic optimization; Chen's 2019 article proposes adaptive charging from measured temperature, distinguished by her predictive rather than reactive modeling; CompetitorX's 2021 product markets faster charging but discloses no technical detail—closing with an explicit statement of the boundary of the inventor's knowledge: "I am not aware of any prior art that combines pulsed charging with real-time predictive thermal modeling as described here." The effective version gives locatable references, explains relevance, articulates the difference, and bounds what the inventor knows. It is also the version that, years later, helps demonstrate good faith on the duty of candor.

Common Mistakes to Avoid

A handful of errors recur often enough to name directly, each the inverse of a lesson above. The first is describing only what you built—a single implementation with no alternatives—which yields a narrow patent a rival sidesteps; describe the concept, not just the execution. The second is omitting the problem and leaping to the solution, which starves the non-obviousness argument and the claim-drafting process of context. The third is marketing language instead of technical language: "revolutionary breakthrough technology" conveys nothing, while "a tenfold reduction in processing latency achieved by caching intermediate results" conveys something a lawyer can claim. The fourth is assuming the attorney already knows your field; even a specialist may not know your specific invention, so define your terms and explain your concepts. The fifth is hiding prior art, a gamble that pays nothing and risks everything, because undisclosed material prior art surfacing later can render the patent unenforceable for inequitable conduct under Therasense. The sixth is waiting too long, letting a publication, sale, or public use ripen into a barring event under Helsinn and § 102 before counsel can file. And the seventh is getting inventorship wrong—omitting a true inventor or naming a non-inventor—either of which can cloud ownership or, if done deceptively, invalidate the patent under the Pannu standard. Notice that every one of these is preventable at the disclosure stage, by the inventor, for free.

Frequently Asked Questions

How long should an invention disclosure be? As long as it needs to be and no longer—but in practice, thorough disclosures run several pages, not several paragraphs. The right test is not word count but completeness: have you described the problem, the mechanism, the alternatives, the advantages, the prior art you know of, the key dates, and the contributors? A complex chemical or software invention may need a great deal of detail; a simple mechanical improvement, less. When in doubt, include it; counsel can trim, but counsel cannot invent facts you left out.

Do I have to have built the invention before disclosing it? No. U.S. law does not require actual reduction to practice before filing—a sufficiently detailed, enabling description constitutes "constructive" reduction to practice (Pfaff v. Wells Electronics, Inc., 525 U.S. 55 (1998)). You should still describe what you have built and tested, and your best mode, but you may and should also describe alternatives you have not built, provided you can describe them in enough detail to enable a skilled person to make them.

Who counts as an inventor? Someone who contributed to the conception of at least one claim—the Pannu v. Iolab Corp. test: a significant contribution to conception, not insignificant in quality against the whole invention, and more than merely explaining well-known concepts. Funders, supervisors, and technicians who merely carried out instructions are usually not inventors. Document everyone's contribution and let counsel make the legal call; inventorship determines ownership, so errors are costly.

What happens if I publish or demonstrate my invention before filing? In the United States you have a one-year grace period from your own public disclosure to file (35 U.S.C. § 102(b)(1)). Miss it, and you lose U.S. rights. Most of the rest of the world has no grace period at all and demands absolute novelty, so any public disclosure before filing forfeits foreign rights. And beware: even a secret commercial sale or offer can trigger the on-sale bar (Helsinn v. Teva, 586 U.S. 123 (2019)). The safe course is to file before any public disclosure or commercial offer—and to tell counsel about any that have already happened so the deadline can be calendared.

Do I really have to disclose prior art that hurts my application? Yes—if it is material and you know of it, the duty of candor under 37 C.F.R. § 1.56 requires it, and the penalty for a deceptive omission is unenforceability of the whole patent (Therasense v. Becton Dickinson, 649 F.3d 1276 (Fed. Cir. 2011)). But you are not the judge of materiality; you list what you know, and counsel decides what to cite to the USPTO. Honest disclosure also strengthens your application, because counsel can distinguish the references up front instead of being ambushed by them later.

Should I patent this at all, or keep it as a trade secret? That depends on whether the innovation can be reverse-engineered, how long you need protection, and whether you can detect infringement. Patents require public disclosure and last about twenty years; trade secrets require secrecy and last as long as the secret holds. A manufacturing process invisible inside a factory may be better as a trade secret; a product feature a competitor can buy and dissect is usually better patented. Counsel can help you weigh it—see our guidance on building a trade secret protection program.

Can an AI tool be an inventor? Under current U.S. law, no—an inventor must be a natural person, and the USPTO and courts have so held. But if an AI system contributed materially to conceiving a claimed feature, that fact matters for both inventorship and enablement, and you should flag it in the disclosure rather than gloss over it. See AI-generated inventions: who owns what the machine creates.

Conclusion: The Disclosure as Foundation

The invention disclosure is the foundation on which the entire patent rests, and like any foundation, its defects are hidden until the structure above starts to crack. A comprehensive, well-organized disclosure lets counsel understand the invention deeply, claim its full scope, support every claim under § 112, satisfy the duty of candor, beat the timing bars, and navigate years of prosecution. A thin one produces a patent that may issue and still protect almost nothing—a certificate suitable for framing and useless for fighting.

When Dr. Marchetti finally sits down to document her fast-charging method, she is not filling out a form for the compliance file. She is writing the source text of a legal instrument her company may rely on for two decades, and she is doing the work that no one downstream can do for her. Describing the problem with precision, explaining the mechanism in detail, documenting alternatives she never built, anchoring her conception date in a witnessed notebook, disclosing the prior art she knows, naming the right inventors, and flagging every date that might trip a bar—none of that is bureaucratic overhead. It is the difference between a patent with teeth and a patent without. Approached that way, the disclosure becomes what it should be: not the least interesting document in the patent process, but the most important one.

Related Articles

• Patent basics and general information concerning patents: a practical guide
• Conducting freedom-to-operate analysis for new products
• Employee invention assignment agreements: drafting for enforceability across jurisdictions
• AI-generated inventions: who owns what the machine creates
• Artificial intelligence and inventorship: global perspectives on machine contributions to innovation
• Responding to patent office actions: strategies for overcoming rejections
• Overcoming obviousness rejections: a comprehensive guide to Section 103 analysis
• Patent eligibility after Alice: strategies for protecting software and business-method innovations
• Inequitable conduct in patent prosecution: navigating the atomic bomb of patent law
• Finding evidence of inequitable conduct in patent prosecution
• Trade secrets in the age of remote work and cloud computing
• Building a trade secret protection program from scratch
• How to license your patent: from valuation to term sheet

Selected Authorities

The authorities below were current as of the date of publication. Patent law and USPTO practice evolve; confirm the current state of the law and consult a qualified patent attorney before relying on any point. This article is general information, not legal advice.

Statutes and Regulations

• 35 U.S.C. § 101 (patent-eligible subject matter; inventorship)
• 35 U.S.C. § 102 (novelty; on-sale and public-use bars; first-inventor-to-file; grace period)
• 35 U.S.C. § 103 (non-obviousness)
• 35 U.S.C. § 112 (written description, enablement, best mode, definiteness)
• 35 U.S.C. §§ 116, 256 (joint inventors; correction of inventorship)
• 35 U.S.C. § 282(b)(3)(A) (best mode eliminated as an invalidity defense, AIA)
• 37 C.F.R. § 1.56 (duty of disclosure / candor to the USPTO)
• 37 C.F.R. § 1.63(c) (inventor's oath and acknowledgment of the duty of candor)

Cases

• Pannu v. Iolab Corp., 155 F.3d 1344 (Fed. Cir. 1998) (joint-inventorship standard)
• Pfaff v. Wells Electronics, Inc., 525 U.S. 55 (1998) ("ready for patenting"; on-sale bar)
• Helsinn Healthcare S.A. v. Teva Pharmaceuticals USA, Inc., 586 U.S. 123 (2019) (secret commercial sale triggers on-sale bar)
• Therasense, Inc. v. Becton, Dickinson & Co., 649 F.3d 1276 (Fed. Cir. 2011) (en banc) (inequitable conduct; but-for materiality and intent)
• Singh v. Brake, 317 F.3d 1334 (Fed. Cir. 2003) (conception; corroboration)
• Dana-Farber Cancer Inst. v. Ono Pharmaceutical Co., 964 F.3d 1365 (Fed. Cir. 2020) (conception does not require proof of success)
• Ariad Pharmaceuticals, Inc. v. Eli Lilly & Co., 598 F.3d 1336 (Fed. Cir. 2010) (en banc) (written-description possession test)
• Amgen Inc. v. Sanofi, 598 U.S. 594 (2023) (enablement of the full claimed scope)
• In re Wands, 858 F.2d 731 (Fed. Cir. 1988) (enablement factors)
• Graham v. John Deere Co., 383 U.S. 1 (1966); KSR Int'l Co. v. Teleflex Inc., 550 U.S. 398 (2007) (non-obviousness)
• Alice Corp. v. CLS Bank Int'l, 573 U.S. 208 (2014); DDR Holdings, LLC v. Hotels.com, L.P., 773 F.3d 1245 (Fed. Cir. 2014) (software eligibility)
• Metallizing Eng'g Co. v. Kenyon Bearing & Auto Parts Co., 153 F.2d 516 (2d Cir. 1946); BASF Corp. v. SNF Holding Co., 955 F.3d 958 (Fed. Cir. 2020) (secret commercial use as public use)

Agency Resources

• USPTO, Manual of Patent Examining Procedure (MPEP), §§ 2138.04 (conception), 2133.03 (on-sale and public-use bars), 2000.01 et seq. (duty of disclosure), 2163–2164 (written description and enablement)
• USPTO, Patent Process Overview, https://www.uspto.gov/patents/basics
• USPTO, Inventor and Entrepreneur Resources, https://www.uspto.gov/learning-and-resources/inventors-entrepreneurs-resources
• USPTO, Patent Pro Bono Program, https://www.uspto.gov/patents/basics/using-legal-services/pro-bono/patent-pro-bono-program

Professional Organizations

• American Intellectual Property Law Association (AIPLA), https://www.aipla.org/
• Intellectual Property Owners Association (IPO), https://ipo.org/
• Licensing Executives Society (USA & Canada), https://www.lesusacanada.org/

For tailored help preparing a disclosure, building lab-notebook discipline, or running a corporate disclosure program, our intellectual property and technology practice works with inventors and R&D teams from first conception through issuance.