Christian's team runs TAM building out of Slack
Christian published the full build for an AI agent that returns verified contacts from a Slack message, wired to 16 data and verification tools. He says running it on one shared Codex account instead of per-person seats saves about $800 a month.
Landing in the inbox
Berman says filters weigh profile picture and day-one volume, not subject line
Alex Berman listed what he says spam filters actually care about, and the subject line is not on it. His six: missing profile picture, blank name field, account age, send volume on day one, links in the first email, and images in the signature. He offers no data behind the list.
Fix those first. Then worry about the copy.
Campaigns and numbers
Christian says Verkada reps book six meetings a day on an MDR model
Christian broke down Verkada's outbound and says its reps book about six meetings a day each, against 1.5 on the old SDR model. He credits an MDR program, a campaign per buyer type, a landing page per prospect, hardware trials and a partner channel. The numbers are his teardown of another company, not a campaign he ran.
Workflows and tooling
Christian's team builds TAM from a Slack message, 16 tools wired in
Christian published the build for an agent that takes a request in Slack and returns verified contacts. It is Hermes from Nous Research on a Hostinger VPS, sitting on one Codex account, connected by API, MCP or CLI to Clay, Icypeas, Prospeo, LeadMagic, MillionVerifier, BounceBan, Firecrawl, Serper, Exa, Apify and others.
We picked Codex over Claude Code because it's cheaper & holds up better when a workflow runs across thousands of rows.