A single LinkedIn post becomes a full engager list: reactions plus comments, pulled with two Apify Actors.
@codyschneider builds it in Claude Code, deploys it to run on a cron, then waterfalls each new engager through enrichment before outreach.
Watch the full agent setup 🤖
last week went on gregs pod
you probably missed it bc you're super busy
here's everything you'll learn
LinkedIn engagement is a hand-raise signal, beats firmographics
Track 10–20 category outliers, get 80% industry coverage
Use your for-you feed, not search, sourcing creators
Business accounts work as lead sources, not just people
Apify API Maestro actors: most stable LinkedIn scrapers
Post reactions plus post comments = full engager list
Daily cron pulls net-new posts, then extracts engagers
ICP-fit research happens before enrichment, not after
Waterfall enrichment: cheapest accurate source first, expensive last
Origami aggregates the entire waterfall behind one call
LeadMagic specifically for mobile phone number enrichment
MillionVerifier before sending, or deliverability dies fast
Buying broker contact data is legal; usage rules differ
Burner domains protect your core domain deliverability
Four domain buckets: cold, marketing, transactional, business
Roughly $200/month total to start sending 10k
Instantly webhooks push positive replies to your agent
Inbox agent answers questions, drives toward booked demos
Program six-month re-touches to revive gone-cold leads
Wire calendar access so agent verifies actual bookings
Agent equals code, thinking loop, live data stream
Don't pay tokens for what cheap CPU does
Agent frameworks are usually bloat for finite problems
Model the human's real process, not autonomous god-box
Organic: interviews, sales calls, Slack, Notion, Gong
"Write good LinkedIn content" produces mid, flaggable slop
Best content is already trapped in internal conversations
Ordinal schedules across accounts, feeds analytics back in
Analytics stream tells agent what to snowball or remix
Repost proven winners every 90 days, never daily
Find what market wants, then build; same for content
LinkedIn paid impressions run ~$22 per thousand — earned media
Topic-based pages work when personal brand isn't appealing
Timestamps
2:28 — Agent one: signal-based cold outbound
9:11 — Apify scraping and extracting engagers
15:47 — Waterfall enrichment, validation, compliance
21:40 — Inbox infrastructure, sending, inbox-managing agent
31:41 — Agent two: organic LinkedIn content engine
もっと見る