VedAI scores every Indian startup on a nine-pillar framework. The pillars are public, the data sources are public, and the cadence is fixed. We don't publish the weights — but we tell you what we use, when we refresh, and how to challenge a score you think is wrong.
Every Indian startup in the IVL universe is assessed on the same nine pillars, each named after one of the Navagraha — the nine celestial bodies of the Hindu astronomical system. We rate each pillar on its own evidence — not in comparison to a "perfect" startup that doesn't exist. The headline VedAI score combines the pillars; the relative weights are intentionally not published, but the pillars themselves are open.
What we don't publish: the relative weights, the model architecture, or the prompt configuration. This protects against gaming and keeps the score directional. What we do publish: every pillar, every input source, and every dispute outcome on request.
If you've looked at a startup profile, you've seen the VedAI /1000 score broken into six buckets — not nine pillars. Both are the same methodology, viewed at two levels. The nine pillars are the measurement layer: how our analysts actually score, dimension by dimension. The six buckets are the reporting layer: how those measurements are grouped for reading a score at a glance.
Why condense nine into six? Because signals overlap. Founder quality, market depth, traction, and business model are all fundamental drivers — so their pillars report together as Core Dimensions. A pillar like Rahu (Innovation & Moat) contributes to more than one bucket: its defensibility signals are core, while its IP portfolio is external validation. Nothing is dropped in the condensation — every pillar's evidence lands in exactly the buckets shown below.
The bucket maximums (450 · 220 · 150 · 100 · 50 · 30 = 1,000) are fixed and public. The pillar-to-bucket routing above is complete — every one of the nine pillars reports into at least one bucket, and no bucket contains evidence from outside the nine pillars.
VedAI does not buy private data. Every pillar is computed from data that is public, regulator-filed, or self-published by the startup. We never use leaked data, scraped private dashboards, or unverified founder claims.
VedAI re-scores the universe nightly. New filings, fresh funding rounds, regulatory updates, and verified news flow in within 24 hours. Pillars that depend on slower-moving signals (governance, compliance) refresh on each batch but typically only change after a real underlying event.
VedAI is a research accelerator — it surfaces patterns, ranks startups for relative comparison, and exposes data gaps. It is not investment advice, a credit rating, or a substitute for due diligence. We mark startups with sparse public data as "insufficient_data" rather than fake a confident score.
If you're a founder, employee, or investor and you believe VedAI has missed a signal — old data, a structural change, a misclassification — write to us. We re-look every flagged score within 7 working days. If we change a score, we publish the change in our methodology change-log. If we don't, we tell you why.
Functional only — no ads, no tracking. Privacy