The Official SaaStr Podcast: SaaS | Founders | Investors SaaStr 845: How SaaStr Built a $5 million Pipeline Machine with 1.5 Humans and 20 AI Agents with SaaStr's Chief AI Officer and Momentum from Salesforce's VP of GTM
19 snips
Mar 11, 2026 
Amelia LeRutte, SaaStr’s Chief AI Officer who scaled from marketing to building multi-agent stacks, explains how she designed 20 AI agents that drove millions in pipeline. She walks through the 3-month deep dive, agent orchestration with Zapier and webhooks, multi-tool workflows that enrich leads and auto-build personalized decks, and the 90/10 buy-versus-build rule.
AI Snips
Chapters
Transcript
Episode notes
Three Month Deep Dive Led To SaaStr Agent Strategy
- Amelia Lerutte locked herself in a room for three months to learn and build AI agents after seeing Delphi's limits.
- That deep dive produced the multi-agent strategy she rolled out for SaaStr leading into the London event.
Agents Create A Slingshot Effect For Small Teams
- Agents multiply output when headcount is constrained, enabling the same team to scale events and sales work.
- Amelia contrasted last year (~10 humans, ~1 agent) with now (many agents) to explain the productivity lift.
Segment Agents By Data Location And Role
- Segment agents by where data lives and funnel role: use Salesforce-native agents for records already in Salesforce and website-focused agents for top-of-funnel visitors.
- Amelia maps AgentForce to Salesforce contacts and Artisan to warm website outreach because each tool has strengths.