← back
The Multi-Agent Architecture That Actually Ships — Luke Alvoeiro, Factory
Takeaway
Long-running multi-agent coding works only with adversarial validation contracts and serial-by-default execution plus structured handoffs — not by parallelizing more agents.
Summary
- Factory's 'Missions' system combines four multi-agent patterns (delegation, creator-verifier, broadcast, negotiation) into a three-role architecture: orchestrator, workers, validators.
- Validation contract is written during planning before any code — hundreds of assertions define 'done' independently of implementation, preventing test-shaped-by-code drift.
- Two validator types: scrutiny (tests/lint/code-review subagents) and user-testing (spawns the app and uses computer-use to click through flows); validators never see the code being judged.
- Workers run serially with internal parallelism only on read-only ops; parallel writers conflict and burn tokens. Longest mission ran 16 days; structured handoffs (commands run, exit codes, issues) self-heal at milestone boundaries.
multi-agentcoding-agentsvalidation
Original description
Everyone's building multi-agent systems, but nobody agrees on how. This talk proposes a taxonomy of five frontier multi-agent strategies and shows what happens when you compose them into a single architecture. Drawing from production data at Factory, we walk through a three-role system (orchestrator, workers, validators) that uses validation contracts, structured agent handoffs, and adversarial verification. We cover the case for serial over parallel execution, why model selection per role is a compounding advantage, and how to design systems that get better with each model generation instead of being made obsolete by them.
Speaker info:
- https://github.com/lukealvoeiro
- https://www.linkedin.com/in/lukealvoeiro
Timestamp:
0:00 Introduction to multi-agent systems and the bottleneck of human attention
1:50 Taxonomy of five frontier multi-agent frameworks
4:04 Introducing 'Missions': The three-role architecture (Orchestrator, Workers, Validators)
6:34 The importance of validation contracts for consistent quality
8:09 Maintaining long-term context through structured handoffs
9:17 The case for serial execution over parallel execution
10:30 Mission control: Monitoring agent progress
11:22 Strategic model selection per role ('Droid whispering')
13:06 Production data analysis: Building a Slack clone
14:34 Designing systems that improve with each model generation
15:51 Conclusion: The shifting economics of software engineering