**The most surprising finding is how lopsided the money is: a few dozen mega-deals make up most of the dollars, and the largest one looks like a data error.** - **Typical deal vs. total dollars.** The median disclosed deal is only about **$1.7M**. The 2,065 deals with a usable amount sum to about **$38.1B** (step-1 and step-8 queries). - **Concentration.** Only **73 deals of $100M or more** (about 3.5% of valued deals) account for **63.6%** of all dollars. Nearly two-thirds of the "Indian startup funding" total comes from a few dozen rounds, so averages and totals say little about a typical startup's raise. - **The largest deal is suspicious.** The top row is **Rapido Bike Taxi, Series B, 27/08/2019, $3.9B**. That one row is **10.2%** of the whole dataset's dollars. It is larger than Flipkart's $2.5B private-equity round (2017) and Paytm's $1.4B and $1.0B rounds. A bike-taxi startup at Series B raising more than Flipkart is almost certainly a unit or entry error, though the data alone can't confirm that. Totals and "share of dollars" figures, including the 63.6% above, are inflated by it. - **Missing amounts are large too.** Of 3,044 rows, only 2,066 have a parsed amount, so **978 (about 32%) have no usable value**. One row, Drums Food, carries a literal `nan` amount, which breaks naive sums. I excluded it from the totals above, which is why those use 2,065 deals rather than 2,066. Any ranking by funding amount is based on only about two-thirds of the deals, and the missing ones may not be random. Because the totals are so skewed, medians or counts by sector and city are more reliable than sums. Before ranking sectors by dollars, check the Rapido row and other outliers above $500M against an outside source.