# SaaStr Books 614 Meetings With Inbound AI Agent > Source: > Published: 2026-06-27 14:10:58+00:00 ### What happened SaaStr founder Jason Lemkin published a conference recap from SaaStr AI Annual 2026 describing the company's inbound AI agent, Amelia, which runs on the Qualified pipeline-management platform. Per the SaaStr post, Amelia handled roughly **2.2 million** website sessions, processed **442,000** individual chats, and booked **614** qualified sponsor meetings for one event, with an average sponsor deal size around **$85K**. Lemkin frames this as replacing the equivalent of 3-10 BDRs, citing high human-rep turnover and limited B-lead follow-up as the problems the agent solves (SaaStr). ### Technical setup Per the SaaStr writeup, Amelia integrates with Salesforce CRM via API, allowing real-time account context lookups, routing logic, and campaign execution. Lemkin emphasizes iterative training - Amelia accumulated roughly 600-1,000 commits over several months, building from a simple contact form replacement into an orchestrated GTM agent. The post also features deployments from Owner.com (83% of new customers start via a free AI product before expanding to a paid plan) and Klaviyo (agents trained on real-time consumer-response feedback as a proprietary moat) (SaaStr). ### Practitioner takeaways Three patterns from the SaaStr writeup are worth tracking as they surface across multiple deployments: First, the highest-ROI agent use case is B leads - prospects with real ICP signal that human reps deprioritize because per-lead expected value is too low. Lemkin claims Artisan (SaaStr's outbound agent) recovered about $500K of revenue from B leads in one year. Second, tight CRM integration is the load-bearing layer - headless API access to Salesforce or HubSpot provides the account context agents need to route accurately without human intervention. Third, DAU/MAU is the wrong quality signal in agentic workflows: every user login implies a task the agent should have handled (SaaStr). ### Caveats All figures above are self-reported by SaaStr in a conference recap; there is no independent third-party audit of the 614-meeting or $85K-ASP claims. The Qualified customer case study corroborates the broad deployment pattern. Practitioners should treat the numbers as directionally useful rather than benchmarked data. ## Key Points - 1SaaStr's Amelia agent (Qualified platform) booked 614 sponsor meetings from 442K chats at one event, replacing multi-BDR inbound teams. - 2The highest-return deployment pattern is pointing agents at B leads - real ICP signal that human reps skip because per-lead value is too low. - 3All figures are vendor self-reported; practitioners should treat them as directional case study data, not benchmarked metrics. ## Scoring Rationale Concrete and useful B2B AI-agent case study with specific numbers from a credible SaaS practitioner (Jason Lemkin/SaaStr). Figures are self-reported and not independently benchmarked, and this is a single-vendor deployment showcase rather than a broader research or platform development. Score reflects solid practitioner relevance at a niche level. Practice interview problems based on real data 1,625 SQL & Python problems across 15 industry datasets — the exact type of data you work with. [Try 250 free problems](/problems)