Three questions stand out. The first is already visible in the data. The other two are possible because of how the table is structured, but I haven't run them yet. ## 1. Is Indian startup funding consolidating into fewer, bigger deals? This is the strongest signal I found. Reported deal counts fall steadily after 2016, while the median disclosed ticket size climbs (query on `raw.startup_funding`, by year): | Year | Deals | Deals with an amount | Median ticket (USD) | |---|---|---|---| | 2015 | 935 | 646 | 1.5M | | 2016 | 993 | 587 | 1.0M | | 2017 | 687 | 456 | 2.25M | | 2018 | 310 | 265 | 4.0M | | 2019 | 111 | 104 | 12.0M | - **What it suggests:** fewer but larger rounds. Deal count drops from 993 in 2016 to 111 in 2019, and the median ticket rises from 1M to 12M. Disclosed totals stay in a similar range: about 8.6B in 2015, 10.4B in 2017, 5.1B in 2018 and 9.7B in 2019. - **Caveats:** - The decline may partly reflect less complete reporting, since the table is a scrape of reported deals. - 2020 has only 7 deals, ending 2020-01-13, so it is a partial year. - Only 2,066 of 3,044 deals (about 68%) have a numeric amount. - The 2016 total came back as `nan`, so some amount values need cleaning before 2016 totals can be trusted. The 2016 median is unaffected. - A few dates are malformed: one parses to year 0015, and 7 are unparseable. ## 2. Where is the money concentrated by vertical and city, and how has that shifted over time? The table has `Industry Vertical`, `SubVertical` and `City Location` (115 distinct cities), plus an amount and a date. That is enough to compare deal count, disclosed dollars and median ticket by sector and city, year by year. It would show whether a few cities and sectors capture most of the capital, and which sectors are gaining or losing share. The 822 distinct vertical labels are messy, so they would need grouping before any ranking is meaningful. This matters for the question of which sectors or cities to target. ## 3. What does a typical raise look like for each round type, and who funds it? - **Round types:** `InvestmentnType` has 57 distinct values, which probably reflect inconsistent labels. Grouping them into seed, angel, Series A/B/C and private equity would give ticket-size benchmarks per stage and sector. - **Investors:** every deal has an `Investors Name`, so you could also measure how often investors repeat, which investors are most active at each stage or in each city, and who co-invests with whom. This would help a founder benchmark how much to raise, and help an investor find peers for a given stage.