Self-improving by default
Every agent swarm evaluates its own output, records what worked, and feeds it forward. The system that ships today is better than the one that shipped yesterday.
About Yex Labs
One studio. Five domains. Every project ships.
Yex Labs LLC runs on self-improving agent swarms, progressive learning cycles, and feedback loops that compound quality across every build. The result: products, public models, and studio systems across archival intelligence, quantitative finance, public policy, personal operations, and AI infrastructure — all built and maintained by a system that gets smarter with every deployment.
Evidence over noise
Measurable performance, calibrated models, and traceable assumptions. No black-box claims — every number has a source.
Tools as leverage
Products remove repetitive bottlenecks so people can focus on judgment, strategy, and craft. AI is infrastructure, not the product.
Studio Systems
The internal layer that lets the studio compound.
These systems are part operating model, part product infrastructure. They are not just portfolio items; they are the leverage layer behind the rest of the catalog.
Studio System
In Development
AI Factory
The orchestration layer behind a one-person studio shipping products, models, and internal systems.
A meta-system that coordinates AI agent workflows, automated content pipelines, knowledge management, and cross-project operations — the reason one founder can ship across five domains simultaneously.
Agent SystemView Details
Studio System
In Development
Progressive Learning OS
A research cycle that turns complex goals into compounding knowledge and shipped outputs.
A tri-track operating system that advances objective execution while compounding learning and producing publishable outputs. Decomposes complex goals into risk-prioritized research, evidence-backed learning cards, versioned decision packets, and day-0 to day-7 execution plans.
Agent SystemView Details
Founder
Chris Schuermyer
Most studios specialize. We built the system that makes specializing optional.
LivesLived turns century-old handwriting into searchable, narrated archives. The Prediction Markets Engine runs quantitative strategies against live event markets. An Oregon tax simulator is changing how policymakers model QSBS provisions. And the AI Factory — the operational backbone — is how one founder ships it all.
What connects these projects isn’t a vertical — it’s a pattern. Each one starts with a messy, unstructured problem that existing tools ignore, and ends with working software built on applied AI, transparent math, and operational rigor. The range is the point.