The Founder Prior-Employer Clearance Standard: Clear, Condition, or Pass
Rebuild each founder's dated timeline from public evidence, score overlap and covenant exposure, and reach the same Clear, Condition, or Pass verdict as your partner.
Key takeaways
- Repo creation dates are recorded by GitHub's servers while commit timestamps are written by the founder's own machine, so any commit older than its repo is the cheapest sign that a build timeline was staged.
- In Refolk's index, 10,165 US profiles carry a Founder or Co-Founder title with "stealth" in the headline, and 24% of a 25-record sample list a placeholder employer, so profile dates alone cannot establish when a founder left.
- The same one-year trailer clause was enforced in Milliken v. Morin (S.C. 2012) and voided in Whitewater v. Alleshouse (Fed. Cir. 2020), so jurisdiction, not clause length, decides trailer exposure.
- Non-solicit pay floors sit below non-compete floors, Illinois at $45,000 against $75,000, so a founder whose non-compete is void can still be barred from recruiting the team they brought along.
- A UK survey of 572 academic founders found most spin-outs take 4 to 12 months to complete, so an option agreement, not a signed license, is the realistic closing condition at term sheet.
- Every finding resolves to Clear, to a named closing condition such as a prior-employer release, a prior-inventions schedule, or a university license, or to Pass, and the team takes the worst founder's verdict.
This standard is for the partner, platform lead, or angel who has to decide before a term sheet goes out whether each founder is actually free of their last employer, university, or lab. It rebuilds each founder's dated timeline from outside-in evidence, scores product overlap and covenant exposure, and ends in one of three verdicts: Clear, Condition, or Pass. For every Condition it names the closing document to send counsel. It is not legal advice. Its job is to tell you what to ask counsel for while asking is still cheap.
What does it mean for a founder to be cleared of a prior employer?
A founder is cleared when nothing they built belongs to a prior employer, university, or lab, and nothing they signed blocks them or the people they brought along. Each founder gets exactly one verdict, and the team gets the worst verdict of any of its founders.
Most published material on this is either a law-firm explainer for founders or a diligence checklist line that says "confirm IP assignment." Neither gives an investor a bar that two partners would grade the same way. This standard fixes the terms, the evidence, and the mapping from finding to verdict.
The terms used throughout:
- PIIA (proprietary information and invention assignment agreement): the document an employee signs that assigns inventions made during employment to the employer.
- Trailer clause (also called a holdover): a PIIA term that extends the assignment to inventions made or disclosed for a fixed period after employment ends.
- Carve-out statute: a state law that limits what an invention-assignment clause can claim. California Labor Code § 2870 is the usual model.
- Non-compete: a covenant that bars the founder from working in a competing business for a set period.
- Non-solicit: a covenant that bars the founder from recruiting former colleagues or customers.
- Institutional IP: a university's or lab's claim on inventions under its own policy, such as Stanford's SU-18 agreement.
- Closing condition: a document that must exist before the round closes, named in the term sheet or on the closing checklist.
| Verdict | What it means | What goes to counsel |
|---|---|---|
| Clear | No finding survives scoring | Standard financing reps only |
| Condition | Every finding is curable by a named document | The list of named documents |
| Pass | A finding cannot be cured before close, or evidence was staged | Nothing. Decline or revisit later |
What can bind a founder
- State lawDecides which contract terms survive: carve-outs, non-compete bans, pay floors
- Employment contractPIIA assignment, trailer clause, non-compete, non-solicit
- InstitutionUniversity or lab policy that vests title through employment or resource use
- The teamEach first hire brings their own agreements, often from the same employer
Which public artifacts date the build, and which can a founder fake?
Six artifacts date a build from the outside: profile end dates, commit timestamps, repo creation dates, domain registration, the incorporation filing, and the first hires' prior employment. Only artifacts recorded by someone other than the founder can anchor the timeline. The rest can only corroborate it.
| Artifact | Who writes the date | How far to trust it |
|---|---|---|
| Profile end date | The founder | Low: edited freely |
| Commit timestamp | The founder's machine | Low: settable before committing |
| Repo creation date | GitHub's servers | High: anchor for commit dates |
| Domain registration | Registrar | Spoofing not assessed here |
| Incorporation filing | The state | Spoofing not assessed here |
| First hires' end dates | Other people | Corroborates or contradicts the founder |
Why repo creation beats commit dates
Git timestamps are written by the client, so they can say anything. Setting GIT_AUTHOR_DATE and GIT_COMMITTER_DATE before committing is enough, because Git does not verify timestamps. GitHub's contribution graph accepts them too, so backdated commits appear on the claimed days. The repo creation date is different: GitHub records it on its own servers. One public demonstration tool works by flagging commits that claim to be older than the repo itself.
That gives you the cheapest staging test in this whole standard. If any commit is older than the repo it lives in, the history was rewritten or imported, and the founder owes you an explanation. A second tell is a run of commits all stamped exactly 12:00:00, because backdating scripts use fixed noon times. I have not assessed how easily domain registration dates, state filings, or package registry dates can be spoofed, so treat them as corroboration only, never as the sole anchor.
Why profile dates cannot settle the timeline
Profile end dates are the weakest artifact. Founders edit them freely, and "stealth" headlines with placeholder employers let one person look both employed and founding from different angles.
| Metric | US | UK |
|---|---|---|
| Total matches | 10,165 | 2,361 |
| Placeholder employer, 25-record sample | 6 (24%) | 8 (32%) |
| Top region in sample | SF Bay / San Francisco (6/25) | London (13/25) |
In Refolk's index of professional profiles, the placeholders read "Stealth Startup", "Stealth AI Startup", or "Stealth Mode". These are small samples and the keyword match is fuzzy. Many sampled records were health companies, so treat the totals as upper bounds and the percentages as a sense of scale, not a rate. The practical point holds anyway: roughly a quarter to a third of sampled stealth founders name no real employer, so you have to date their departure from someone else's record.
The best someone-else record is the first hires. If three engineers left the same employer as the founder, their end dates bracket the founder's. Finding them by hand means searching name by name. Refolk returns the people who left a given employer and joined a given startup from one plain-English query, and that cross-check is hard for a placeholder employer to survive.
The order of a clean timeline
- Last day at employerCorroborated by press, first hires' end dates, or founder documents
- Repo or org creationRecorded by GitHub's servers; the anchor for commit dates
- First commitMust not predate repo creation
- Domain and incorporationRegistration and filing dates, used as corroboration
- First hires joinTheir last days at a shared employer feed the non-solicit check
How does state law decide invention-assignment and trailer-clause exposure?
The state where the founder was employed decides whether an invention-assignment clause can reach a side project and whether a trailer clause survives. The forum matters more than the clause's length.
Carve-out states
Sources disagree on which states have a carve-out statute, so record which list you used.
- One source: California plus Delaware, Illinois, Minnesota, North Carolina, Washington, and Nevada. It notes that Nevada takes a more employer-favorable approach.
- A 50-state survey: 10 jurisdictions: California, Delaware, Illinois, Kansas, Minnesota, New Jersey, New York, North Carolina, Utah, and Washington.
Where a carve-out applies, it excludes inventions "developed entirely on your own time, without employer equipment, supplies, facilities, or trade secret information, that do not relate to the employer's business or anticipated research and development and do not result from work you performed for the employer." Every clause of that test has to hold. The relatedness clause is where most false Clears come from: if the invention relates to the employer's "demonstrably anticipated" research, California does not protect it.
Two more limits matter:
- Contractors get nothing. None of the state carve-outs apply to independent contractors. A founder who was a consultant at the prior company is graded on the consulting agreement alone.
- Notice is a question for counsel, not a win. California Labor Code § 2872 requires an employer that includes an invention-assignment clause to attach or reference the § 2870 text. A secondary source says California, Illinois, Kansas, Minnesota, and Washington all require written notice, and that an assignment given without it could be unenforceable. I could not establish the case law on what actually happens when notice is missing.
Trailer clauses
Every documented trailer-clause case I found involved a one-year trailer. Longer ranges get quoted, but I have not seen them established publicly, and nobody has published how often each length appears.
| Case | Year | Forum | Outcome |
|---|---|---|---|
| Ingersoll-Rand v. Ciavatta | 1988 | N.J. Sup. Ct. | Not enforced |
| Applied Materials v. AMEC | 2009 | Cal. | Void |
| Milliken v. Morin | 2012 | S.C. Sup. Ct. | Enforced |
| Whitewater v. Alleshouse | 2020 | Fed. Cir. | Void (Cal. 16600) |
In Applied Materials, the court held invalid a clause requiring assignment of inventions described in patent applications or disclosed to third parties within one year after termination. In Milliken, the court called a one-year holdover "eminently reasonable." Whitewater voided one as a restraint on the profession under California's Section 16600. Washington is unsettled: no published Washington decision has squarely decided whether a post-employment trailer is enforceable.
The same one-year trailer was enforced in South Carolina and voided under California law. Forum decides, not length.
When grading, record the date the trailer window closes. If the window is still open at term sheet and the forum is not one where trailers have been voided, that is a finding.
How do non-compete and non-solicit floors apply to a founder?
State law governs. Four states void nearly all employee non-competes, others enforce them only above a pay floor, and non-solicit floors sit below non-compete floors. A founder who is free to compete can therefore still be barred from recruiting the team.
The federal backstop is gone. The FTC voted 3-1 to dismiss its appeals and then removed its non-compete rule (Federal Register document 2026-02866). It still pursues case-by-case enforcement, as its order against Rollins, Inc. shows. For founder diligence, that means state rules:
- Void: California, Minnesota, North Dakota, and Oklahoma treat nearly all employee non-competes as void.
- Scheduled void: in Washington, almost all non-competes become void regardless of pay from 30 June 2027.
- Pay floor: elsewhere, a non-compete or non-solicit is enforceable only if the employee earned above a set floor.
| State | Non-compete floor | Non-solicit floor | Note |
|---|---|---|---|
| Washington | $126,858.83 | Not established | Void from 30 Jun 2027 |
| Colorado | $130,014 | $78,008.40 | |
| Oregon | $119,541 | Not established | 12-month maximum |
| Virginia | $78,364.52 | Not established | |
| Illinois | $75,000 | $45,000 | |
| Maine | $63,840 | Not established |
These are the 2026 schedule figures. Floors are indexed and step up on fixed dates, so re-check them before every use. Score each agreement against the floor in force when it was signed, which may differ from today's number.
Tag each covenant with exactly one of these:
- Void: the state bans it, or the founder's salary was below the floor in force.
- At risk: the floor for that covenant type is not established, the founder's salary is unknown, or the employer's state and the founder's state apply different rules.
- Enforceable: the state enforces it, the founder's salary was above the floor, and the covenant term has not expired.
When the team came from the same employer
This is the case partners most often miss. A founder's non-compete may be void while their non-solicit is enforceable, because non-solicit floors are lower: Illinois sets $45,000 against $75,000, and Colorado $78,008.40 against $130,014. If the first three engineers left the same employer within weeks of the founder, the question changes from "can the founder compete" to "did the founder recruit them in breach." List every early hire's prior employer and last day, and run the non-solicit check for each founder who overlapped with them. Refolk will turn up this pattern across a portfolio's pipeline, not just one deal.
What changes for academic founders and spinouts?
For an academic founder the question is whether the university's policy vests title in the work. If it does, the question becomes whether a license or option will be in hand by close. The test is usually employment duties plus how much the founder used university resources.
I have checked one policy in detail, Stanford's. Do not assume other universities match it. Under Stanford's policy:
- All faculty, graduate students, postdoctoral fellows, staff, and other employees, including student and temporary employees, must sign the Stanford University Patent and Copyright Agreement, known as SU-18.
- Title vests in Stanford for inventions made in the course of university responsibilities or with more than incidental use of university resources.
- Stanford does not claim IP from students' consulting or other work outside their academic programs, provided there was no more than incidental use of resources.
- SU-18 does not apply to inventions that qualify fully under California Labor Code § 2870.
For any other university, read the policy text and record two things: who must sign it, and where the resource-use threshold sits. Using lab space or other specialized facilities counts as more than incidental use.
License timing becomes closing timing
Published figures on how long a license takes vary:
- A UK independent review surveyed 572 academic founders. Most spin-outs took 4 to 12 months from the point the founder and the tech transfer office agreed to spin out to completing every process.
- One US investor's spinout playbook reports that 2 or 3 months is not uncommon at top schools (Stanford, MIT, Harvard, Northwestern), but that the same offices can take up to 12.
- AUTM, the tech transfer professionals' association, does not report a single standard figure.
The clearance procedure, step by step
Seven steps take a founding team from raw artifacts to a verdict. The first four take about four and a half hours per founder, split between an analyst and the partner.
Founder prior-employer clearance
- Collect dated artifactsPull each founder's profile end dates, GitHub org and repo creation dates, commit history, domain registration, incorporation filing, and the first hires' prior employers and last days (about 2 hours per founder). Done is a dated list with a source and a "recorded by" value for every row.
- Build the timelineLine up last day at the employer, repo creation, first commit, domain, incorporation, and first hires in date order (about 1 hour). Done is every date before the last day flagged, and every commit older than its repo flagged as possible staging.
- Classify jurisdictionRecord the employer's state, the founder's state, employee or contractor status, any carve-out statute, whether notice was given, and the trailer length read against the four one-year trailer cases (about 30 minutes). Done is a forum and a trailer window close date per founder.
- Score restrictive covenantsApply the state ban or the floor in force at signing to each founder's salary, for the non-compete and separately for the non-solicit where colleagues came along (about 1 hour). Done is each covenant tagged void, at risk, or enforceable.
- Check institutionsFor academic founders, name the governing policy, decide whether resource use went beyond incidental, and record license status. Done is license in hand, option in hand, or neither.
- Score product overlapScore from 0 to 3 whether the product relates to the employer's business or demonstrably anticipated research, citing the evidence. Done is a score a second partner can reproduce from the cited sources.
- Issue verdict and conditionsAny Pass trigger gives Pass, any curable finding gives a named Condition, and otherwise the founder is Clear. Done is a verdict per founder, a team verdict equal to the worst one, and each condition mapped to a disclosure-schedule item.
The overlap score
Overlap gets scored separately from the timeline because it decides whether a carve-out applies at all.
| Score | Test | Typical evidence |
|---|---|---|
| 0 | Unrelated to employer's business or research | Different market, buyer, and technology |
| 1 | Adjacent: shared buyer or stack, different problem | Same customer, unrelated product |
| 2 | Relates to employer's business or anticipated research | Employer ships, publishes, or researches in the area |
| 3 | Same problem the founder worked on there | Founder's prior title or team names the area |
The verdict rules
A founder gets Pass if any of these is true:
- A commit predates its repo's creation, or timestamps fall at uniform noon, and the founder cannot explain it with documents.
- Overlap is 3, server-recorded build artifacts predate the last day, and no prior-employer release can be obtained before close.
- A non-compete tagged enforceable is still running and covers the product, with no release.
- An academic founder used resources beyond incidental use, and the institution will not grant a license or option before close.
A founder gets Condition if there is no Pass trigger and any of these is true:
- Overlap is 2 or 3 and the build started after the last day.
- Any artifact predates the last day and overlap is 0 or 1.
- A trailer window is still open in a forum that has not voided trailers.
- The carve-out depends on statutory notice that cannot be shown.
- The founder was a contractor at the prior company.
- A non-solicit is tagged enforceable or at risk, and colleagues came along.
- An academic founder needs a license or option.
Otherwise the founder is Clear.
Timeline against overlap
Founder: Prior employer / institution: Employer state / founder state: Employment type (employee / contractor / student / postdoc): Date | Artifact | Source | Recorded by (founder / third party / server) | Before last day? (Y/N) ____ | Last day at employer | | | ____ | GitHub org / repo creation | | server | ____ | First commit | | founder | ____ | Domain registration | | | ____ | Incorporation filing | | | ____ | First hire 1 joins (prior employer: ____, last day: ____) | | | Commit older than repo? (Y/N): Uniform 12:00:00 timestamps? (Y/N): Carve-out statute (list used): ____ Notice shown? (Y/N): Trailer clause length: ____ Forum: ____ Window closes: ____ Non-compete: floor in force at signing ____ / salary ____ / tag (void, at risk, enforceable) Non-solicit: floor in force at signing ____ / colleagues brought ____ / tag Institution: policy ____ / beyond incidental use (Y/N) / license, option, or neither Overlap score (0 to 3): ____ Evidence: ____ Findings: Verdict (Clear / Condition / Pass): Conditions:
One sheet per founder; the second partner grades from the same rows without seeing the first partner's verdict.
How does founder clearance go wrong?
Clearance fails when a partner trusts an artifact the founder controls, or a legal protection that does not actually apply to this founder. Each failure below produces a false Clear, and each has a check you can run in minutes.
- Backdated commits. How it lies: the history makes the build look like it started after the founder left. Check: is the first commit earlier than the repo's creation date? Are the timestamps all exactly 12:00:00?
- Edited profile end dates. How it lies: the founder moves their end date, or lists a placeholder employer that hides the conflict. Placeholders turned up in 24% of a US sample and 32% of a UK sample in Refolk's index. Check: cross-reference press releases and the first hires' end dates.
- Assuming the carve-out. How it lies: a California founder is marked Clear because of the state, while the product sits in the employer's demonstrably anticipated research. Check: grade from the overlap score in step 6, not from the state.
- Treating missing notice as a win. How it lies: the employer's PIIA omitted the statutory notice, so the partner assumes the assignment fails. The only support for that is a secondary source. Check: make it a Condition for counsel, never a Clear.
- Overlooking contractor status. How it lies: the founder describes "working with" the prior company, and the partner applies employee carve-outs. The statutes do not protect contractors. Check: confirm the engagement type before applying any carve-out.
- Accepting an academic "side project". How it lies: the founder says the work was done outside the lab, but lab space or specialized facilities were used, which is more than incidental use. Check: search for co-authored papers and grant acknowledgements that describe the same work.
- Missing the team non-solicit. How it lies: the founder's non-compete is void, so the covenant box is ticked, while a lower-floor non-solicit still binds them. Check: tag the non-solicit separately for every founder whose former colleagues joined.
Two further limits apply to the standard itself. I have not assessed whether domain, state-filing, Form D, or package-registry dates can be spoofed, so none of them should be the only anchor for a date. And the legal reads above come from a small number of cases and surveys. Wherever a finding turns on an unsettled question, such as a Washington trailer or a missing notice, the verdict is Condition, and the question goes to counsel.
How do findings become closing conditions for counsel?
Each finding maps to one named document, and each document maps to a disclosure-schedule item in the purchase agreement. The partner sends counsel a list of documents, not a list of worries.
| Finding | Closing condition |
|---|---|
| Overlap 2 to 3, built after last day | Prior-employer release |
| Pre-departure work, overlap 0 to 1 | Prior-inventions schedule plus carve-out analysis |
| Open trailer window, enforcing forum | Prior-employer release covering the window |
| Beyond-incidental university resource use | University license or option |
| Enforceable or at-risk non-solicit, colleagues hired | Waiver or counsel's opinion |
| Missing notice or contractor status | Counsel review of the prior agreement |
These conditions feed the company's representations. In a priced round, the company represents that none of its former or current employees have excluded any inventions; one source points to Section 2.19 of the model NVCA Stock Purchase Agreement. Purchase agreements filed with the SEC carry company reps on PIIAs and on conflicts with former employers, qualified by "to the Company's knowledge." I have not established the current NVCA text or whether it includes founder-level reps, so counsel should read the current model directly. That knowledge qualifier is exactly why the work happens before term sheet: a rep only covers what the company knew, and your timeline is how you find out what it should have known.
Subject: [Company] founder prior-employer clearance: closing conditions Team verdict: Condition (worst founder verdict) Founder: [Name] Verdict: Condition Finding 1: [e.g. Repo created [date], [n] days before last day at [Employer]; overlap score 1] Condition 1: Prior-inventions schedule listing [project], plus carve-out analysis under [state statute] Finding 2: [e.g. Non-solicit tagged enforceable; [n] engineers joined from [Employer] within 90 days] Condition 2: Waiver from [Employer] or written opinion that the non-solicit does not reach these hires Disclosure schedule: map each condition to the IP and former-employer reps in the SPA Open questions for counsel: - [e.g. PIIA lacks statutory notice; enforceability not established] - [e.g. Washington trailer clause; no published decision on point] Evidence: worksheet attached, every row sourced and dated
Send once per deal; one block per founder with a Condition, citing the worksheet rows.
Before the verdict leaves the partnership
- Every founder has a worksheet with a source and "recorded by" value on every dated row
- No commit is older than its repo's creation date, or the founder has explained it in writing
- Each founder's last day is corroborated by someone other than the founder
- Employee or contractor status is confirmed before any carve-out is applied
- Every non-compete and non-solicit is tagged against the floor in force at signing
- Each academic founder is marked license in hand, option in hand, or neither
- The overlap score cites evidence a second partner can open
- A second partner graded the same rows independently and reached the same verdict
- Each Condition names one document and maps to a disclosure-schedule item
How do I keep this standard current?
Re-check the parts that move on a schedule before each use, and re-score any founder whose dates or salary inputs changed between term sheet and close. The standard's logic is stable. Its inputs are not.
- Pay floors. Several are indexed annually and step up on fixed dates, some of them mid-year. Illinois, for example, rises to $80,000 for non-competes and $47,500 for non-solicits on January 1, 2027, and Tennessee introduced a $70,000 floor from July 1, 2026. Before each use, check the source for every state in Dataset A, and keep the floor in force at signing next to today's.
- Bans and scope changes. Track scheduled dates such as Washington's move to near-total voidness, and sector rules that take effect mid-year.
- Federal enforcement. With the rule removed, watch for FTC case-by-case orders rather than a rule.
- Carve-out lists. Re-read the 50-state survey and note which list you graded against.
- University policies. Re-read the policy text for each institution you see, not just the ones you have seen before.
- Financing documents. Have counsel confirm the current NVCA model's IP and former-employer reps.
Run the procedure at term sheet, then again in the last week before close. That second pass is short. It catches a trailer window still open, an option that never arrived, or a new hire from the founder's old team.
Questions practitioners ask
Is a founder still employed while building a startup an automatic pass?
No. Building before the last day is a finding, not a verdict. It becomes a Condition, usually a prior-inventions schedule plus carve-out analysis, when the founder was an employee in a carve-out state and the work relates neither to the employer's business nor its demonstrably anticipated research. It becomes a Pass when overlap is high, server-recorded artifacts predate departure, and no prior-employer release can be obtained before close.
Can I trust GitHub commit history to show when a founder started building?
Not on its own. Git does not verify timestamps. Setting GIT_AUTHOR_DATE and GIT_COMMITTER_DATE before committing is enough, and GitHub's contribution graph shows backdated commits on the claimed days. Anchor on the repo creation date, which GitHub records on its servers, and flag any commit that claims to be older than the repo. Uniform 12:00:00 timestamps are a second tell, because backdating scripts use fixed noon times.
Does California Labor Code 2870 protect a California founder's side project?
Only if the project passes every part of the test: built entirely on the founder's own time, without employer equipment, supplies, facilities, or trade secrets, unrelated to the employer's business or demonstrably anticipated research, and not the result of work done for the employer. Relatedness is where false Clears happen. The statute also covers employees only, so a founder who was a contractor or consultant gets no carve-out.
How long does a university spinout license take, and should I wait for it before closing?
Usually longer than the round. A UK review surveying 572 academic founders found most spin-outs take 4 to 12 months from agreement with the tech transfer office to completion. One US investor's playbook reports 2 or 3 months at top schools, but up to 12 at the same offices. Make an option agreement the closing condition, and let counsel decide how the full license is handled after close.
What is the risk if the founding team was hired away from the same employer?
The risk is a non-solicit, which can bind a founder even when their non-compete is void. Where states set both floors, the non-solicit floor is lower, for example Illinois at $45,000 against $75,000. List each first hire's prior employer and last day, check each founder's non-solicit separately, and where one is enforceable make a waiver or counsel's opinion the closing condition.
Did removing the FTC non-compete rule change founder diligence?
It removed the federal rule, so state law decides again. The FTC removed its non-compete rule by Federal Register document 2026-02866, but it still enforces case by case, as its order against Rollins, Inc. shows. Score each founder against their state. California, Minnesota, North Dakota, and Oklahoma void nearly all employee non-competes, and other states enforce them only above pay floors that move on fixed dates.
Try it on the search you came here for
Stop building boolean strings. Just describe the person.
Type one sentence. I plan the search, read GitHub, public LinkedIn and Crunchbase records, and the open web as it is right now, and hand back a ranked list with the reason next to every name.
01Describe them
One plain sentence. Role, city, stack, stage, whatever matters to you.
02I read the web live
GitHub, public LinkedIn and Crunchbase records, the open web. Not a database that went stale last quarter.
03You read the shortlist
Ranked, with the reasoning under every name. Open a profile, ask a follow-up, narrow it down.
- Staff backend engineers in NYC who shipped Rust in production
- Series A fintechs in SF under 50 people, growing headcount this year
- Maintainers of fast-growing Rust web frameworks on GitHub
- No boolean, no filters, no seat to buy. One box.
- Read at search time, so a profile updated yesterday counts today.
- Every step visible as it runs, every name with its reason.
500 free credits on sign-up. No card, no demo call. See real searches.