Osparna asks one question at a time: is this specific company, in this specific deal, worth the risk. Oluwadi widens that to a whole network: who's building and funding what, region by region, thesis by thesis, across the venture map. Both are, in the end, questions about companies — the innermost and middle rings of the same onion.
Ayni is the outer ring. Its question is different, and it doesn't start with a company at all: what is this place, actually — apart from whether it produces fundable startups.
A region is not its startup density
Read a place only through its venture activity and you get a real answer to a narrow question. You don't get the place. A county can have zero active startups and still have 40,000 residents, a hospital system, a flood risk score, a police department, a graduating class, a set of election returns that tell you something true about how it governs itself. None of that shows up in a cap table. All of it shapes what actually happens there.
Startups per capita is the report we shipped first, and it's still the sharpest single number Ayni produces — but it's a ratio, and a ratio needs both sides to mean something. The numerator (real companies, read the same way Oluwadi reads them) only means something next to a denominator that's honest about the place itself: how many people actually live there.
So the datasets aren't a startup-adjacent afterthought. Population by county, all 50 states plus DC and every territory ACS5 doesn't reach — the reference layer everything else gets divided by. Monthly climate summaries, temperature and precipitation and snow, county by county. The nationwide law-enforcement agency roster. County-level health measure estimates. A composite natural-hazard risk score. Presidential returns by county, every cycle. Employment and wages by county and industry and quarter. Home prices, job postings, electricity costs, commuting patterns — who works where, who lives where, and whether those are the same county at all. National laboratories, interstate compacts, the convening institutes and policy think tanks and advocacy organizations that shape a place without ever appearing in a funding round.
None of that is about startups. All of it is about the place a startup would have to be built in.
Not all of it is a number, either. Eight times a year the Federal Reserve publishes the Beige Book — plain-language write-ups from each of the twelve Districts, built from conversations with local bankers, business contacts, and community leaders about how things actually feel on the ground: hiring, prices, confidence, worry. In a consumption-driven economy, that feeling isn't separate from the market — it's the engine of it. Consumer sentiment doesn't just describe conditions, it moves them, months before a GDP number or a jobs report catches up. Reading how a place feels is reading where its market is headed, not reading around it.
Sacred exchange, made visible
Ayni is a Quechua word — reciprocity, mutual obligation, the practice of giving with the expectation that what you give will someday be returned, not necessarily by the same person or in the same form. It's not a transaction. It's a relationship to a place and the people in it, held over time.
You can't practice that with a place you've only read one axis of. An investor who knows a region's startup density but not its flood risk, its crime-agency coverage, its actual population under 18, is not seeing the place — they're seeing the part of it that was already legible to their existing tools. Reading the whole place, on its own terms, before you decide what to do with what you found, is itself a form of reciprocity. It comes before the ask.
That's the outer ring's job. Osparna evaluates the deal. Oluwadi maps the network the deal sits inside. Ayni holds the ground both of them are standing on — the weather, the population, the risk, the institutions, the history of who actually voted and who actually got sick and who actually has a job — independent of whether any of it is fundable this quarter.
We're early. The catalog will grow before it's finished growing, and every dataset is on the Data page with its own last-update timestamp, honestly, because an honest empty shelf is more useful than a full one you can't trust. Start with a place you think you know. Ask it the wider question.