AI Agents Seattle August 2026
Adam Jacob cut one code review from 4.6 million tokens across 23 agents to 506,000 across 3 by replacing the LLM coordinator with deterministic code. Casey Margell spent nine months chewing glass in production and concluded the biggest risk with AI coding agents is efficiently building the wrong thing. One is building a factory; the other is surviving the field. On August 11 they compare notes. AI Agents Seattle is a meetup for engineers actively building with AI agents. It exists because the space keeps changing faster than anyone's map, and the people building in it need to share hard-won lessons rather than hot takes. This August, two speakers coming at the same problem from opposite starting points, one building an agentic factory from scratch, the other adapting agents into a running business, will talk through their trade-offs, their failures, and what they're changing next. How to build a software factory. Adam Jacob, CEO and Co-Founder, Swamp Club Adam Jacob builds agentic systems as repeatable, auditable software factories, and his agents are generating what he calls shockingly consistent architecture. He'll share how he cut a complex code review workload from 4.6 million tokens across 23 agents to 506,000 tokens across 3 by moving coordination out of the LLM and into deterministic code, and why he thinks it is too early to bet on frameworks when adaptive building blocks are winning. Adventures in Adapting: nine months chewing glass building AI agents. Two very different approaches, the trade-offs I made, and how I evaluate what's next in a space that just keeps changing. Casey Margell, Co-Founder and President/COO, Indisea Software Casey Margell has spent nine months building AI agents in production, most recently wiring Claude into Indisea's Slack and connecting it to Jira and Confluence; it filed 9 correctly formatted Jira tickets from a PDF of meeting notes on day one. He'll share why he thinks the biggest risk with AI coding agents is efficiently building the wrong thing, and the product-owner-agent pattern he uses to keep agents aligned with requirements after he steps away.
