The sample is in the result blocks above: 15 rows pulled across the dataset's date range, and a second set of 15 evenly spaced by Sr No (100, 200, … 3000). The second set is the more representative one. Each row is one reported funding deal, keyed by `Sr No`, with the startup, date, industry, city, investors, round type, amount and remarks. A few things stand out in the sample that matter if you analyse this data: - **Amounts are text in Indian digit grouping.** Values such as `1,70,00,000` mean 17,000,000, and `N/A` appears often (e.g. Digit, EvenGini, Faircent). The column needs cleaning before it can be summed or averaged. - **The column layout shifts for older rows.** In the 2015 rows (e.g. Razorpay, Lendingkart, Sulekha.com), `Industry Vertical` holds a free-text description, `SubVertical` is `nan`, and `Remarks` carries the round label (Series A, Series C, Bridge funding). In 2016–2019 rows, `Industry Vertical` is a cleaner category (Consumer Internet, Logistics, eCommerce) and `SubVertical` holds the description. - **Category labels are inconsistent.** The same industry appears as "E-commerce", "ECommerce" and "eCommerce". The same city appears as "Bangalore" and "Bengaluru". - **`InvestmentnType` is coarse in this sample.** It is mostly "Private Equity" or "Seed Funding", even for deals the remarks call Series A or Series C. The remarks field is needed to see the actual stage. - **Missing values are written as the string `nan`.** This shows up in `Remarks` and in some city and vertical fields (e.g. Faircent has no city or industry). - **Investors are one comma-separated string per deal.** Some entries have trailing commas (e.g. "Accel Partners, "), so splitting them needs care. The sample covers deals from early 2015 to mid-2018, across cities such as Bangalore, Mumbai, Gurgaon, Chennai, New Delhi and Noida. I can profile null rates, clean the amounts to numbers, or break deals down by vertical, city or year next.