Startup
it does a lot of thingsfintechSeedIndia1 million usdWants from investors: money
Workflow
Describe the startup
by You · Free
Name, what it does, sector, stage, geography, round size and what you want from investors. Checked and saved.
Saved.
Write the scoring rubric
by Gemini Flash · 1 Gemini call
Gemini reads the profile and proposes the weights, the thresholds, what each topic class means for this startup (and the highest relevance each may have) and the keywords that decide investor fit. Code then clamps every number: weights rescale to 100, thresholds are ordered, caps cannot rise. If Gemini fails, a plain default built from the profile is used and labelled. You can edit it.
Written by Gemini Flash from the profile, then checked and clamped by code. Review it below.
Choose who can be scored
by Code + database · Free
The shared list of investors. Only investors that have had a completed evidence search are scored; the rest are left out and counted, never shown as quiet. Demo accounts are excluded.
Gather the evidence
by Code + database · No search credits
The verified signals from the last 30 days already collected for those investors: what each investor itself did, with its source link and date. No new search runs, and nothing from your other startups or your private notes is read.
Judge each signal for this startup
by Gemini Flash · About 1 Gemini call per investor with evidence
For each investor, Gemini reads their signals and says how relevant each is to this startup and which topic class it falls in. Answers are validated; an omitted or invalid answer leaves the signal out, and the investor is marked not conclusive. Judgements are saved, so re-scoring is free.
Score every investor
by Code · Free
A fixed formula, no model: recency, relevance (capped by topic class), evidence strength, investor fit and relationship, using this rubric's weights. Gates stop “reach out now” when the lead signal is weak, off-topic, low-confidence or social-only.
Read the shortlist
by You · Free
Investors grouped by recommended action, each with its lead evidence, its score breakdown and why. Nothing is sent from this app: you decide who to contact.
Appears once the evaluation has run.
The evaluation reads evidence that is already collected. Gemini judges each signal against this startup (about 30 to 60 short calls, cached, so a second run costs none), then the scores are calculated by fixed rules. It does not search the web, so it uses no Tavily credits.
Rubric
Relevance is weighed slightly higher to prioritize investors actively funding Indian fintech startups, while keeping relationship low for an early-stage cold raise.
Standard thresholds are used to ensure timely outreach when strong signal alignment occurs for a Seed round.
Fit terms capture seed and pre-seed stage investors focusing on Indian financial technology and adjacent B2B software.
Sector terms: fintech, financial services, payments, lending, neobank, insurtech, wealthtech, banking, paytech · Geography: india, south asia, asia
The sector focuses specifically on fintech and financial services localized to the Indian market.
Shortlist
Not evaluated yet