Recently Funded Companies — Rounds, Amounts & Investors

If you sell to startups, a funding announcement is the best two weeks of the year to reach them. The money is in the bank, the plan is being written, and nobody has spent it yet. The problem was never finding out that someone raised — it is getting a clean, deduplicated, machine-readable list of who raised, how much, from whom, and who inside the company now has budget — fast enough to act on it.

That is this Actor. Give it a stage, a sector or nothing at all, and it returns one row per company and round: the amount everyone agrees on, the investors, the announcement date, the company's domain, the roles that just got budget, and how warm the trigger still is. Re-run it and it tells you only what is new.

---

Who it is for

• Outbound and SDR teams — a daily list of accounts that just raised, with the buyer roles to target and the announcement to reference in the first line of the email.
• AI / GTM agents — one call, a flat dataset, no login and no key. Every row is already a decision: signalScore, signalTags, sellTo, buyingWindow.
• VCs and corp dev — deal flow by stage, sector and geography, with the syndicate on each round.
• Recruiters — a Series A or B with hiring-wave-likely is a company about to open ten roles.
• Data teams — a clean funding feed to join onto your CRM, refreshed on a schedule, deduplicated for you.

---

What makes it different

	Typical funding scraper	This Actor
Row shape	one row per article	one row per company + round, merged across every outlet that covered it
Amount	whatever the one article said	the figure the reports agree on (a bigger number in one headline is usually total capital raised, not this round)
Investors	often missing	lead investor + the syndicate, cleaned and deduplicated
Domain	rarely	resolved for most companies, so you can enrich straight away
Re-runs	the same rows again	only what is new since your last run, plus follow-on rounds and revised amounts
Output	raw data	raw data plus the signal: tags, who to sell to, the buying window, a 0–100 score

---

What you get per company

{
  "company": "Lightfield",
  "domain": "lightfield.ai",
  "round": "series-a",
  "amountUsd": 47000000,
  "amountText": "$47 Million",
  "valuationUsd": null,
  "announcedDate": "2026-09-10",
  "daysSinceAnnouncement": 1,
  "leadInvestor": "a16z",
  "investors": ["a16z", "Y Combinator"],
  "industry": "AI",
  "hqLocation": "San Francisco",
  "description": "Lightfield is building a CRM for companies that run on AI agents.",
  "signalScore": 86,
  "signalTags": ["fresh-capital", "just-raised-series-a", "announced-today", "large-round",
                 "hiring-wave-likely", "tier1-lead-investor", "new-since-last-run", "ai"],
  "sellTo": ["CEO", "VP Sales", "Head of Marketing", "Head of People", "Head of Engineering"],
  "buyingWindow": "Reach out now — inside the 14-day window when a just-raised company answers most",
  "isNewSinceLastRun": true,
  "isFollowOnRound": false,
  "previousRounds": [],
  "mentions": 4,
  "sourceUrl": "https://…",
  "sourceUrls": ["https://…", "https://…"]
}

Three dataset views are built in: Recently funded companies (the signal table), Outreach (domain, who to sell to, buying window, what they do) and New since last run (the delta).

---

Use it in three ways

1 · Zero-config. Run it with nothing set: every funding round announced in the last 7 days, worldwide, sorted by signal strength.

2 · Your ICP. Pick the stages you sell into, add a sector word or a country, set a minimum round size:

{ "roundTypes": ["series-a", "series-b"], "keywords": ["fintech"], "locations": ["Germany"],
  "minAmountUsd": 5000000, "daysBack": 14 }

3 · A scheduled feed of what changed. Set a watchId, turn on onlyNewSinceLastRun, and schedule it daily. The first run stores the baseline; every run after it returns only the rounds you have not seen — which is exactly what an agent should act on.

{ "watchId": "fintech-series-b", "roundTypes": ["series-b"], "onlyNewSinceLastRun": true, "daysBack": 7 }

Watching specific accounts instead? Put them in companies (names or domains) and you get a row only when one of them raises.

---

How the signal is built

• Merge, then report. The same raise is announced by five outlets with five different headlines. They are merged into one signal — with a mentions count, every source URL, and the amount the reports agree on. Wide coverage is itself a signal, and it is on the row.
• Remember, then diff. Each watch keeps a private memory of every company and round it has already reported. That is what powers isNewSinceLastRun, isFollowOnRound, previousRounds and amountChangedFrom — the delta, not the dump.
• Score, then sort. signalScore (0–100) combines recency, round size, stage, breadth of coverage and investor quality, so the top of the dataset is where to start calling.

---

Chain it

Then run	To get
Decision-Maker Finder	the actual people in sellTo at each funded company, with work e-mails
Hiring Intent Scraper	which of them turned the money into open roles — the second wave of the signal
Technology Lookup	the stack they are buying into
Company Finder	more companies like the ones that just raised

---

Notes

• No login, no cookies, no API key. Everything read is public.
• No other paid Actor is ever started — the run costs what it reads.
• Amounts in other currencies are converted at an indicative rate so filters and sorting work worldwide; the amount as reported is kept in amountText with its own currency.
• Sources are public announcements, so a company that never announced its round will not appear.

Pricing

Pay per event: you pay per delivered company row. Duplicate reports of the same round, filtered-out rounds and empty runs are never charged. See PRICING.md.