Funding Rounds Scraper & News Monitor

Collect company funding announcements from public news. Get one deduplicated row per funding round, with company details, investors, amounts and source evidence. Use the archive to filter rounds, resume pages, or return only new records for a monitor.

The default sources are PR Newswire, Tech.eu, Entrackr and Disrupt Africa. Coverage depends on their current public listings and access rules. This Actor builds its own archive. It does not include a proprietary funding database or complete historical coverage.

Quick start

1. Copy this input into Apify Console and start the Actor.
2. Open Funding rounds in Output, or open the default dataset.
3. Export the dataset as JSON, CSV or Excel. Check SUMMARY for source failures and coverage limits.

{
  "maxItems": 10,
  "maxArticles": 14,
  "maxArticlesPerSource": 5,
  "enrichCompanyWebsites": false
}

Rules mode needs no provider key. maxItems limits returned matching rounds; it does not guarantee that many discoveries. The Actor saves collected rounds before it returns a filtered page.

Use cases

• Find recent company funding announcements for investment research or sales prospecting.
• Filter an existing archive by date, funding stage, amount, country or known company attributes.
• Schedule repeated runs with onlyNew: true to deliver new IDs for a monitor.

Inputs

Field	Default and behavior
maxItems	100; 1–10,000 matching archived rounds per page. The spending limit can reduce this.
sources	PR Newswire, Tech.eu, Entrackr, Disrupt Africa. [] disables discovery. Optional Wamda, FinSMEs and EU-Startups may be blocked.
startUrls	[]; up to 500 public article URLs, collected alongside the selected sources.
maxArticles	40; 1–500 article attempts per fresh run. Discovery, FX and website requests are additional. A small budget can reduce article work.
maxArticlesPerSource	15; 1–100 candidates per feed or listing. No deep archive crawl.
requestDelaySeconds	1; 0.5–60 seconds per host, increased for robots crawl-delay. One request at a time; no access bypass or automatic retry.
since	Optional ISO date/time. Filters the announcement calendar date. Rules use publication date as a proxy.
round	[] means all. Examples: seed, series-a, debt, grant, unknown.
minAmountUsd, maxAmountUsd	Optional nonnegative amounts in USD. When set, missing or unconverted amounts are excluded.
countries	[]; ISO-2 codes or English country names. unknown includes missing country. Geography from rules is a heuristic.
industryGroups, verticals, naicsCode	Optional exact labels or NAICS prefix. Rules leave these unclassified; use unknown to include missing values.
employeeBuckets	Optional bands: 1-10, 11-50, 51-200, 201-500, 501-1000, 1001-5000, 5001-10000, 10001+, unknown.
enrichedOnly	false; requires domain, country and industry group. Rules alone normally cannot satisfy this.
minimumConfidence	0.7; range 0–1. A heuristic score, not a probability of correctness.
enrichCompanyWebsites, maxEnrichments	true, 8; optional company-homepage JSON-LD, capped at 100 websites. Does not scrape LinkedIn.
convertCurrencies, fxRates	true, {}. Historical Frankfurter rates where available. Override with USD per currency unit, for example {"INR":0.01055}. Missing rates leave USD amounts null.
onlyNew, monitorId, resetMonitor	false, optional ID, false. Delivery history is separate for each monitor. Without since, only-new uses a rolling 30-day window. Reset clears delivery history, not the archive.
cursor	Optional saved NEXT_CURSOR. Reuse the same filters, monitor and state store. Expires after seven days; continuation reads the saved snapshot without collecting news.
stateStoreName	independent-funding-rounds-v1. Named key-value store for your persistent archive. Run only one writer per store. Choose a new name for an independent archive.
maxStoredRecords	10,000; range 100–50,000. Stops at the cap without silently deleting records. Export and use a new store for larger archives.
extractionMode, openaiApiKey, llmModel	rules, no key, gpt-4.1-mini. Optional llm mode requires your OpenAI key. Provider fees are separate. Live LLM accuracy is not verified for this release.

Omit optional fields instead of passing null. An empty filter list means no filter. Input validation occurs before collection. Invalid input writes INPUT_ERROR and SUMMARY.outcome=INPUT_REJECTED; for compatibility, the cloud run then ends successfully with no results.

Outputs

The default dataset contains funding rounds. It contains no diagnostic or error rows. Each row includes:

Fields	Meaning
id	Stable round identity. Use it to deduplicate downstream.
company	Stable company ID, name, domain, description, headquarters, industry, employee band and other available attributes. Missing facts are null or empty lists.
round, investors	Normalized stage and identified investors. An unknown stage is unknown.
amountUsd, amountUsdCents	USD amount in whole units and integer hundredths. Null when unavailable.
amountOriginalCurrency, amountOriginalCents	Original currency and integer hundredths, including INR. Divide cents by 100 for whole units.
announcedAt, discoveredAt	Announcement date or publication-date proxy, and first discovery timestamp.
_meta	Article links, supporting quotes, extraction method, confidence, FX source, inferred fields and warnings.

For example, a verified CureBay article states Rs 100 Cr. Its extracted amount fields are:

{
  "amountOriginalCurrency": "INR",
  "amountOriginalCents": 100000000000,
  "amountUsd": 10550000,
  "amountUsdCents": 1055000000
}

The historical conversion used 0.01055 USD per INR. Rates and evidence are retained in _meta.

The default key-value store contains SUMMARY, NEXT_CURSOR and OUTPUT. SUMMARY.outcome distinguishes SUCCEEDED, PARTIAL, EMPTY, BUDGET_EXHAUSTED, INPUT_REJECTED and FAILED. Partial source failures retain useful results. If all requested collection fails, the run fails and retains its summary. Empty results alone do not indicate complete coverage.

Use https://api.apify.com/v2/datasets/DATASET_ID/items for complete JSON. Add ?format=csv for CSV. Private datasets require authentication. The Output links use the current run's dataset and storage IDs.

Pricing and limits

The configured price is $60 per 1,000 saved funding rounds, plus $0.001 per run start at 256–1,024 MB. Platform usage is included. One result event applies to each funding round saved in the default dataset. The Actor-start event also applies to empty, invalid-input and failed runs. Empty results, rejected inputs and failed articles produce no result events. A failed or partial run can still contain paid results already saved before failure.

Returning the same archived round in a later run is another delivered result. Use onlyNew: true with a stable monitorId to avoid normal repeat delivery. A crash between dataset storage and monitor checkpoint can still repeat a row; downstream delivery is at least once. Cursor-only runs also charge for the results they deliver.

For 10 saved rounds, event charges total $0.601. For 100 rounds, they total $6.001. A run with no saved rounds costs only the $0.001 start event. Use a spending limit of at least $0.061 for a start and one result. The pricing configuration permits $0.001, but Apify can reject limits below one result's price before starting a run. These limits do not guarantee discovery.

Set a finite maximum charge and timeout for each run. A budget that cannot fund one result stops collection before source or provider calls. Optional LLM calls use your key and can incur provider charges even if an article yields no matching result. The Actor never replaces an invalid supplied key with an owner-funded key.

Public-source access, robots rules and extraction quality can change. No full-history, universal coverage or verified-contact-data guarantee applies. Review source evidence before making decisions. Monitoring and pagination require the same persistent store. Records from earlier versions are not automatically repaired; use a new store to recollect them.

FAQ

How do I monitor new rounds?

Add "onlyNew": true and a stable "monitorId": "my-funding-monitor". Reuse the same stateStoreName. Schedule the Actor in Apify Console. One writer per store avoids archive conflicts.

How do I query the archive without collecting more news?

Use "sources": [] and "startUrls": [], plus your filters and the same state store. Follow NEXT_CURSOR with the original filters to retrieve further pages.

Why are some fields null?

Rules only retain facts that the extraction logic can identify. Industry groups, NAICS and LinkedIn numeric IDs are not inferred in rules mode. Optional website enrichment only reads public structured company data.

Why did I get no results?

Check SUMMARY. Filters may exclude all rounds, no matching announcement may be available, or source access may fail. onlyNew also removes already delivered IDs. All-source access failure is reported as a failed run.

Operational telemetry

When enabled by the operator, Sentry records one bounded input snapshot and original error stacks with safe operation targets. Credentials and URL authentication are masked. Article bodies, generated answers, HTTP payloads and stack locals are excluded.