Hey Product Hunt! 👋 I built EjectSeat because I got caught off guard by a layoff at a previous company — and when I went back and looked at their 10-K filing from 6 months earlier, every signal was right there in plain English. I'd just never read it. So I built a tool that reads it for you. Here's what EjectSeat actually does under the hood: 1. It pulls the full SEC evidence bundle for any US-listed company — 10-K, 10-Q, all 8-K events (not just the filtered ones), and NT filings 2. It scrapes Motley Fool earnings call transcripts, specifically the Q&A section — because prepared remarks are scripted, but CEOs go off-script in Q&A and that's where the real signals live 3. A single Claude Sonnet call reads the full bundle and returns a structured risk assessment with a CLEAR / WATCH / LIKELY / ACTIVE state and a 0–90 score 4. A post-validator strips any unattributed claims so you're only seeing evidence that traces back to a real source The score that drove v7 was a Snap bug I found in v6 — the pipeline was filtering 8-K filings by item number, which meant it missed the actual restructuring disclosure. Fixed by reading full text and letting the AI decide what's material. Happy to answer questions about the architecture, the signal design, or how you'd use this for your situation. And if you try it — run your own company first. The results are often surprising. → ejectseat.io

EjectSeat

Know your layoff risk before the Slack message arrives

Know your layoff risk before the Slack message arrives