01
signal
Funding, hiring, pricing, and prod launches in one view
We stitch SEC filings, job boards, changelog feeds, Crunchbase deltas, GitHub activity, and pricing-page diffs into a single company record that updates on its own.
v2026.04 · soft launch
Signals,
not narratives.
AI Startup Insights is the briefing desk for operators who need to know what the AI ecosystem is actually doing, not what it tweeted about. 2,471 companies tracked. Every four hours. Structured. Queryable. Quiet.
MONDAY BRIEF / WK 15
Helical Labs raised $28M Series A
Obvious Inc. posted first enterprise AE
Stacklight removed free tier
Kernelbase shipped SOC 2 Type II
2,471
Companies tracked
18,300
Signals ingested / week
62
Source feeds
Every 4h
Dashboard updated
Sample feed, last 72h
What the feed looks
on a Tuesday
FUNDING 2026-04-14 09:18
Helical Labs
Series A closed at $28M. Lead: Redpoint.
HIRING 2026-04-14 07:02
Obvious Inc.
First enterprise AE posted. Signal: moving upmarket.
PRICING 2026-04-13 21:40
Stacklight
Removed free tier. Starter now $49/mo.
LAUNCH 2026-04-13 16:55
Kernelbase
Shipped SOC 2 Type II + dedicated deployments.
SHUTDOWN 2026-04-12 11:21
Brightwater AI
Landing page now returns 410. Team dispersed.
MODEL 2026-04-11 14:03
Cerenote
Switched default from GPT-4 to their own fine-tune.
This is a structured, redacted sample. Full feed is gated.
Product
Four things we do.
Nothing else.
02
synthesis
Weekly movement reports, not five-page newsletters
Every Monday: what shifted, who hired, who stopped. One page. Links back to raw sources. Read in four minutes.
03
filter
Slice by stage, vertical, geo, infra stack
Want only US-based vertical AI companies between Series A and B that added a head of sales in the last 30 days? That query runs.
04
api
API first, dashboard second
Every record, every signal, every diff is available as a JSON endpoint. Build your own watchlists, your own alerts, your own internal briefings.
Mechanics
Three steps.
That is the whole stack.
- 01
We crawl, dedupe, and timestamp every public signal.
- 02
An LLM layer clusters noise into structured facts.
- 03
You get a daily brief, a live dashboard, and an API.
Why this exists
The AI ecosystem produces more noise per hour than any sector in recent memory. Twitter threads, Discord leaks, funding rumors, model announcements that turn out to be wrapper products.
Operators we talked to, VCs, product leads, founders staring across at competitors, spent ten to fifteen hours a week trying to piece together what was actually happening. Most of that time was spent on the same three searches done differently.
So we built the thing they kept manually doing. It runs now. You can have it.
Questions we get
Honest answers,
nothing clever.
01 Is this a newsletter? +
No. The briefing is a byproduct. The core product is a structured database of AI companies with their signals over time. Some people live in the dashboard. Others prefer the Monday email. Both work.
02 Where does the data come from? +
Public sources only: SEC filings, Crunchbase deltas, LinkedIn and Wellfound job boards, GitHub public repos, pricing pages, changelogs, podcast transcripts, and press releases. No scraping of login-walled data.
03 How often is the dashboard updated? +
Every four hours for high-frequency signals (pricing, job posts). Every 24 hours for filings and funding. Manual review daily on anything the model flags as ambiguous.
04 Can I get custom alerts? +
Yes, on the Team plan. Define a query in natural language (for example: any vertical AI company in healthcare hiring a compliance lead) and we deliver a Slack or email ping.
05 Do you cover companies outside the US? +
Yes. Roughly 38% of tracked companies are outside North America. Europe and Singapore are strongest. China coverage is thinner and we are honest about that.
Access is gated
Ask for a seat.
We reply in 48 hours.
We let new accounts in slowly. It keeps the data quality high and our costs in check. No waitlist theater, just a form and a real reply from one of seven humans.
Request access