The rows already pulled are a fairly representative sample of the dataset. Step 1 took every 200th `Sr No` (200 through 3000), which spreads the sample across 2015–2018. Step 0 holds 15 other rows from a separate query, so it covers different deals. **What the sample shows** - **Grain:** each row is one funding deal, with startup, date, city, investor(s), round type and amount. Examples are Shop 101 (Mumbai, Private Equity, 10/07/2018) and Samunnati (Chennai, Private Equity, 10/01/2018). - **Industry mix:** Consumer Internet, eCommerce, Finance, Logistics, Technology and Auto appear most often among the 2016–2018 rows. - **Round types:** only a few values appear, mainly "Private Equity" and "Seed Funding". The card lists about 66 distinct `InvestmentnType` values, so the rest are rarer or messier variants. - **Cities:** Bangalore, Mumbai, Chennai, Gurgaon, New Delhi, Noida and a few smaller cities such as Bhopal and Ahmedabad. **Data quality to know before analysing** - **Amounts are text.** `Amount in USD` is a VARCHAR with Indian digit grouping, for example `2,35,00,000` for 23.5 million, and `N/A` for undisclosed amounts. Several sampled rows are `N/A`, including The Gourmet Jar, EvenGini and Faircent. Strip the commas and cast before summing. - **Missing values are the string `nan`.** This affects `SubVertical`, `Remarks` and sometimes `City Location`. Faircent has `nan` for industry, subvertical and city. - **The schema shifts over time.** From about mid-2015 backward, `Industry Vertical` holds a long free-text description such as "Online Finance lending platform", `SubVertical` is `nan`, and `Remarks` holds the real round label such as "Series A". From 2016 onward the verticals are cleaner categories with a separate subvertical. - **Inconsistent spellings:** "Bangalore" and "Bengaluru" both appear, as do "ECommerce" and "eCommerce". - **Investor lists are free text.** They are comma-joined, sometimes with a trailing comma (for example "Accel Partners,") or "and" separators. - **Column naming:** the date column is named `Date dd/mm/yyyy` and the city column `City Location`, with a double space. The sample shows the data needs cleaning before vertical, city or round-size analysis. The main steps are casting amounts, normalising city and vertical names, and unifying the pre-2016 and post-2016 layouts.