From 25 to 32: Inside the Signal Expansion That's Redefining B2B Intelligence

TL;DR: 7 new signal types in 5 weeks. SEC Form D private funding, federal contracts, Product Hunt launches, podcast appearances, Hacker News, conference & CFP signals, and government RFPs. All live in the Signal API today.

Five weeks ago, Autobound’s signal engine covered 25 data sources. Today it covers 32.

Each one is a new pipeline — new source, new parsing, new entity resolution, new enrichment. Seven categories of buyer intelligence that didn’t exist in our system a month ago, all pulling from publicly available data that most teams aren’t monitoring.

We built these because customers kept asking: “Can you tell me when a company raises money before it hits Crunchbase?” and “Can you tell me when a company wins a federal contract?”

Now we can. Here’s what we shipped.

SEC Form D — Private Funding

When a company raises private capital — venture round, convertible note, private placement — they file a Form D with the SEC under Regulation D. This happens weeks before any press release or Crunchbase update.

We parse the full XML from EDGAR’s EFTS feed twice daily. The filing contains how much they’re raising, how much has been sold, investor count, exemption type, directors and executives named on the filing, industry group, and revenue range. We extract all of it.

{
  "signal_type": "sec-form-d-funding",
  "company_name": "Acme AI Inc.",
  "company_domain": "acmeai.com",
  "data": {
    "total_offering_amount": 12000000,
    "total_amount_sold": 9500000,
    "total_investors": 8,
    "exemptions": ["Rule 506(c)"],
    "industry_group": "Technology",
    "revenue_range": "$1M - $5M",
    "related_persons": [
      {"name": "Jane Chen", "roles": ["Director"]},
      {"name": "Mark Torres", "roles": ["Executive Officer"]}
    ],
    "summary": "Acme AI filed Form D for a $12M offering under Rule 506(c)..."
  }
}

The timing advantage is real. Form D filings are public record, updated daily, and almost nobody in sales intelligence is watching them systematically.

Federal Contract Awards

Every federal contract and grant above $100K, pulled daily from USASpending.gov, filtered to tech and services NAICS codes.

The hard part here is entity resolution. Government records use legal entity names that don’t match anything in your CRM. We resolve recipients to company domains via SAM.gov registration data. You get contract value, awarding agency, NAICS classification, performance period, and place of performance — all structured.

RFP & Government Opportunities

The demand side. While contract awards tell you who already won, the RFP feed from SAM.gov shows who’s about to compete.

Full solicitation lifecycle — presolicitations, active RFPs, sources sought, combined synopses. Each includes NAICS classification, estimated value, response deadline, set-aside type, and point of contact.

Product Hunt Launches

We hit the PH GraphQL API twice daily, aligned to their batch release schedule. Product name, tagline, description, vote and comment counts, topic tags, maker profiles. A company launching on Product Hunt is showing you their technical priorities in real time.

Podcast Appearances

When a CEO goes on a podcast, they talk candidly for 30-60 minutes about direction, challenges, and priorities. Almost nobody is capturing this systematically.

We index episodes via PodcastIndex.org, identify ones featuring executives from companies we monitor, and run audio through OpenAI Whisper for transcription. The enrichment layer extracts pain points (with intensity scores), strategic initiatives (with urgency scores), technologies mentioned (with adoption status — evaluating, using, migrating from), and competitors referenced.

Hacker News

Three subtypes from the Algolia HN API:

• Show HN launches — companies announcing products to the most technical audience on the internet
• Trending stories — companies being discussed with real sentiment from actual users
• Who’s Hiring threads — monthly hiring data with role descriptions and tech stack mentions

Conference & CFP Signals

Aggregated from confs.tech and CallingAllPapers — event details, topic tags, CFP deadlines, speaker info. A company sponsoring KubeCon or submitting a talk to a data engineering summit is telling you what they’re investing in.

The Processing Layer

Building a scraper is the easy part. Any engineering team can hit the EDGAR API in an afternoon. The hard part is everything after:

• Entity resolution — Form D lists “Acme AI, Inc.” Your CRM has “acmeai.com.” We bridge that across every source. 99%+ domain coverage.
• 300+ tag taxonomy — Not just “funding” but exemption type, investor count range, offering size bucket, industry classification. Filterable out of the box.
• LLM enrichment — Summary, pain points with intensity scores, initiatives with urgency scores, technology mentions with adoption status, competitor references.
• Schema normalization — Whether the signal came from EDGAR, Product Hunt, or a podcast transcript, the output schema is identical. Integrate once, every new signal type just works.
• Deduplication — Same funding round surfaces in a Form D, a news article, and an exec’s LinkedIn post. We dedup across sources.

We also support semantic search via vector embeddings — query signals in natural language, not just structured filters.

What’s Next

We shipped 7 in 5 weeks because the infrastructure work is done — processing layer, entity resolution, schema normalization. Each new source builds on that instead of starting from scratch. We’ve evaluated 900+ public data sources across 49 research documents. The pipeline is deep.

Beyond new signal types: webhook delivery, portal improvements, real-time enrichment APIs, expanded geographic coverage. More soon.

Try Them

All 7 are live. Existing customers: they’re already flowing. Add the new signal_type filters to your API calls.

New here: browse the Signal Directory (400+ subtypes across 32 types), read the docs, or book a demo.

— Daniel