AI for GTM — Cerebras (2026-04-20)
Summary
- anthony-landsberg is an AE at cerebras (AI chip/inference compute, 7-9 figure contracts); previously 5 years at AWS. Niv led the interview.
- Primary GTM stack: ample-market (CRM + LinkedIn signals, auto-sequences, dialer) + Claude Code for research, copy, persona analysis, podcast summarization.
- AI multiplied outreach volume 4-5x (from ~100/day to 400-500/day) but conversion rate stayed flat — volume is solved; conversion/hit-rate is the unsolved problem.
- Biggest pain: getting through to people (spam filters, iOS call screening). LinkedIn response 2-3%, email <1%, cold call 3-5%.
- Voice agents: has NOT used them. Still prefers phone (“old school”). Long tail of smaller Cerebras customers is a self-service tier he sees automation potential in.
- Cerebras filed S-1 on Friday (IPO pending).
Decisions raised
(None — GTM research conversation, no commitments.)
Action items raised
(None from this session.)
Open questions raised
(None new.)
Entities
- anthony-landsberg — primary research subject; AE at Cerebras
- cerebras — his employer (AI chip/inference compute, pre-IPO)
- ample-market — his primary outbound GTM tool
- niv — led the interview
- saar-arbel, vitali, nizan-shifman, ben — Harmony attendees
Concepts
- cross-funnel-intelligence — Ample Market’s CRM+LinkedIn signal unification resonates with Harmony thesis
- agent-in-call — not raised; Anthony prefers traditional phone; useful contrast
- self-learning-voice-agent — indirect signal: voice agents not yet adopted by experienced AE
Notable quotes
“We went from about 100 outreach a day to 400-500. But the conversion rate stays exactly the same.” — Anthony Landsberg
“What’s missing is AI to get us better hit rates. Getting through to people is the number one problem.” — Anthony Landsberg