**The time dimension is the announcement month, January to May 2022, with one table per month.** The `Founded` column is a company founding year, not a funding timeline. The notable breaks are an April spike in dollar value and a May collapse in volume and data quality. **1. Round counts decline, then fall sharply in May.** - Rows per month run 115 (Jan), 96 (Feb), 98 (Mar), 95 (Apr), then 42 (May), a drop of about 56% from April. - May is probably a partial or incompletely captured month rather than a real market collapse. The card notes that some May rows are misaligned or blank. **2. April has a very large funding spike.** - Parsed USD totals are about $4.10B (Jan), $3.49B (Feb) and $2.75B (Mar). April jumps to about $45.08B, roughly 11 to 16 times the earlier months, then falls to about $0.30B in May. - The spike comes from a few extreme rows, not broad strength. The top five April rows are Accel ($18.3B), Ola ($5B), Mahindra Logistics ($2.5B), Verse Innovation ($1.7B) and ShareChat ($1.4B). Accel alone is about 40% of the month. - Accel is an investor, not a startup, and its stage is "undisclosed", so that row looks like a data error or a fund-size figure. I'd treat the April total with suspicion. - The April median is $30M, versus $5.45M (Jan), $6M (Feb), $12M (Mar) and $5.25M (May). April's typical round is therefore also larger, so the spike is not purely one outlier. March's median is also elevated relative to Jan and Feb. - The largest single round in each of the other months is $700M (Jan), $400M (Feb), $350M (Mar) and $67.6M (May). **3. Undisclosed and unparseable amounts rise over time.** - Rounds without a clean dollar amount are 7 (Jan), 10 (Feb), 19 (Mar), 15 (Apr) and 22 (May). - In May, 22 of 42 rows (about 52%) have no usable dollar amount, and 21 of them have a blank Amount. That is why the May dollar total is not comparable with the other months. **4. Format breaks in the Amount field.** - January has a misaligned text value ("Vaibhav Anant") in Amount. - May has a bare "6400000" with no `$`. - January has an Indian-style comma pattern ("$1,40,000"), which could be misparsed. - Amount is stored as text and is inconsistent across months, so month-to-month totals depend on how it is cleaned. **Takeaway:** the real signals are the May drop in volume and the April jump in round size. Both are heavily affected by data quality, so compare months using the median or excluding outliers rather than totals.