A community tutorial published on dev.to walks through building a daily feed of newly funded US startups from SEC Form D filings, using a third-party Apify actor called the Form D Scraper. The author, posting under the handle agents001code, frames the piece around a practical problem: press releases cover only a fraction of funding rounds, and paid databases are expensive and often lag. The tutorial is a how-to, not an announcement of a new SEC service or a verified product review, so its claims should be read as the author's description of their own tool.
What the post outlines is a mechanism that is both public and official. A short notice known as Form D goes to the SEC whenever a US company sells securities privately under Regulation D; it is due within 15 days of the first sale of securities. Form D/A is the amendment form, used for instance when an offering extends beyond a year or its details change. On EDGAR, the SEC's filing system, each filing appears and carries the company name and business address, an industry group, the total offering amount, the amount already sold and remaining, the date of first sale, the security type, the claimed exemption such as Rule 506(b), and the number of investors so far.
Format, not access, is the obstacle the author names. One filing at a time is how EDGAR presents material, and the useful fields sit embedded in XML, so bulk monitoring is awkward without tooling. EDGAR full-text search and the daily index are what the actor described in the post reads; each filing is opened, and one row per filing with consistent columns is returned. The tool is independent, the post explicitly notes, and is not affiliated with or endorsed by the SEC.
The walkthrough covers three usage paths. First, a manual run in the Apify Console with input fields for days back (default three), form types (D and D/A by default, with 8-K and 8-K/A optionally added), US state codes, industry group keywords, a switch to exclude pooled investment funds, a minimum offering amount, a record cap, and an SEC User-Agent string. The author says a default run of 50 records takes roughly 15 seconds and can be downloaded as CSV, Excel or JSON.
Second, the post describes scheduling. Turning on an "only new since last run" option makes the actor remember the accession number of every filing it has delivered, stored in a named key-value store in the user's Apify account, so later runs skip those filings and the user is charged only for new ones. The author recommends keeping the lookback at three days so late filings and weekend gaps are still caught without duplicates, and suggests a cron schedule of 0 7 * * 1-5 in the America/New_York timezone. One caveat the author flags: the memory is tied to the chosen filters, so changing a filter produces a fresh memory unless a state store name is set manually.
Third, the tutorial shows a Python path using the official apify-client package, installing it with pip and supplying an API token from the Console's Integrations section. The sample code starts a run with a spend cap, iterates the resulting dataset, and writes selected fields to a CSV. The author notes that the call returns None if the run fails, so the check matters, and that iterate_items() pages through the dataset automatically.
The post includes a sample output record for a company called Alercell Inc, filed 2026-10-19 in Montana, in the Biotechnology industry group, with a total offering amount of $1,500,000, $35,000 sold, $1,465,000 remaining, a first sale date of 2026-10-12, equity securities, a Rule 506(b) exemption, and three investors. The author explains several field semantics: totalOfferingAmount is null with an indefinite flag when the issuer filed the amount as Indefinite, which the post says is common for funds; comparing amount sold against remaining indicates whether a round is mostly closed or still open; and amendments marked as D/A can show a round that has grown.
On privacy, the author states that executive and director names are off by default and that the switch to include them should only be used where there is a lawful basis to process that personal data. This is a notable caution for anyone building prospect lists from the data.
Pricing is described as pay-per-event at $6 per 1,000 filings saved, or $0.006 per filing, with filings dropped by filters not charged. The author offers two illustrative budgets: a filtered feed returning about 40 new filings a day costs roughly $0.24 daily, or about $5 over 21 working days, while an unfiltered feed of every new Form D, which the post puts at a few hundred records on a typical day, would run around $1.80 a day at 300 records. The author advises checking the Store page for current pricing before budgeting, and notes a maximum total charge can be set per run.
The post suggests downstream uses: pushing rows into a CRM such as HubSpot via Zapier, Make or a webhook with the accession number as a unique key to avoid duplicates; connecting the dataset to Google Sheets for a shared morning list; posting filings above a size threshold to Slack with the filing URL; and running longer date ranges, which the author says go back to 2009, to count raises by industry, state and size over time.
Gaps are something the author addresses candidly. Up to 15 days after the first sale, money has already started coming in before Form D, a notice, is filed, so a round already announced publicly will rarely be first reported by the feed. Late filings happen at some companies, and some raise without filing Form D at all, so the author characterizes the feed as a strong source with gaps. Investors and valuations are not listed on Form D. Addresses and phone numbers appear as filed and may be stale or belong to a rarely used office. No filings appear on weekends or US federal holidays, US issuers only are covered by the data, and contact details are company-level with no personal emails.
For freelancers, designers and developers who sell to startups, recruit for them or invest in them, the practical appeal is timing: a company that has just closed a round has fresh budget and new hiring plans. The tutorial's contribution is a documented, repeatable pipeline from a free official source to a filtered daily list, with deduplication and a spend cap. The tradeoff is dependence on a third-party paid actor layered over public data, plus the inherent lag and coverage gaps of Form D itself.
What remains unverified is the actor's reliability, accuracy and current pricing, since the post is the author's own description of their tool and no independent testing is reported. The sample record's 2026 dates are presented as illustrative output rather than a confirmed filing. Readers weighing this approach should treat the post as a starting recipe and validate output against EDGAR directly before relying on it for outreach or investment decisions.