**The most counterintuitive finding is that the "funding landscape" is mostly a handful of mega-deals, and the largest one looks like a data error rather than a real deal.** - **The median deal is tiny.** Across the 2,065 deals with a usable amount, the median is about **$1.7M**, yet total reported funding is about **$38.06B** (step 1). The typical startup deal and the headline dollar total describe very different markets. - **A few deals carry most of the money.** Only **73 deals of $100M or more** (about 3.5% of valued deals) account for **63.6% of all dollars** (step 1). Averages or totals by city or vertical will mostly reflect whichever sector or city hosted a few large rounds. - **The biggest deal looks wrong.** The top row is **Rapido Bike Taxi, Series B, 27/08/2019, $3.9B**, about **10.2% of all dollars** by itself. It is larger than Flipkart's biggest round ($2.5B, Aug 2017) and Paytm's ($1.4B) (step 0). A bike-taxi startup out-raising Flipkart is implausible, so this is probably a unit or scale error (for example $3.9M recorded as $3.9B). I can't confirm that from the data, but I would verify it before using any sector or city totals. - **One bad row broke my later totals.** A row for **Drums Food** (21/07/2016) has an amount that parses to `NaN`, not a number (step 0). That is why my later queries returned `nan` for totals and counted **74** deals over $100M instead of the clean **73** (steps 2–7). Any naive sum of this column will be poisoned the same way unless that row is filtered out. - **A third of deals have no amount.** About **978 of 3,044 rows (32%)** have no usable amount (step 7). Any "who raises the most" ranking silently ignores them. **Takeaway:** treat dollar totals in this dataset as fragile. Use medians and deal counts for benchmarking, and exclude or check the top outliers (starting with Rapido) before drawing sector, city or trend conclusions.