About the founder
One person, working from the filings. Not a data vendor, not a team, not a fund.
vincii is built and run by a single developer — a University of California, Berkeley Regents’ and Chancellor’s Scholar, the most prestigious scholarship awarded by the university to the top 1% of undergraduates.
That is here because people ask, and because a yearly subscription is partly a bet on the thing still being here next year. It is not the reason to trust the figures. Every number on this site names the filing it came from and links to it at sec.gov, so you can check any of them without knowing or caring who computed it. A credential is a claim. A citation is a check. This site runs on the second kind.
What one person means
- Nothing is decided by committee. A wrong number reported in the morning is a fix the same day, because there is nobody to escalate it to.
- There is no sales team, which is why the prices are on a page instead of behind a form, and why the free tier is the whole method rather than a teaser.
- Support is email, answered by the person who wrote the code. Not a queue, but also not around the clock.
- Your work is portable. Every page cites sec.gov, and Analyst plans and above export everything to a file. Nothing here is a format only this site can read.
Where the numbers stop
The methodology sets out what is computed, what is withheld, and why. The short version: figures come from SEC filings and nowhere else, and where a filing does not support a figure the page says so rather than printing a zero. Nothing here is advice, and nothing here is a rating.
For investors
Self-funded to date. vincii is open to conversation about equity. The case is structural rather than promotional, so here it is in the same form as everything else on this site — figures, and where they come from.
- No data licence, and none needed. Every figure is derived from SEC filings, which are public domain. Competitors in this category pay for market data; the largest recurring cost in the sector is one this product does not carry. Deliberately no share prices — the one input that would require a licence is computed in the reader’s browser against a price they type.
- Marginal cost per subscriber is close to zero. The site is statically generated and served from a CDN, with a single edge worker for entitlement. Adding a subscriber adds no compute; adding a company adds one file. The economics do not change between a hundred subscribers and a hundred thousand.
- The dataset is the barrier. 5,303 companies computed from filings, 803 with published pages, twelve measures each and nineteen investors’ published methods mapped onto them. The mapping between the two is the part that took the longest and the part that cannot be scraped.
- It compounds without attention. Filings are checked three times each business day and pages recompute themselves. The corpus grows and freshens whether or not anyone is at the keyboard.
For product questions, pricing or support, use the contact page instead — this address is for investment enquiries only.