About Zetesis AI
The name Zetesis (ζήτησις) originates from classical Greek philosophy, signifying relentless, methodical inquiry and the unyielding pursuit of empirical truth. In an era of rapid AI proliferation, we adhere to the scientific ethos: engineering transparent, controllable, and deterministic foundations for autonomous intelligence.
Our Vision
We envision a future where complex software engineering, scientific discovery, and operational workflows are accomplished through seamless collaboration between human experts and autonomous agent collectives. By developing modular orchestration protocols and verifiable state machines, we give creators true determinism over agent behavior.
Engineering Principles
We reject treating black-box outputs as production-grade truth. Every decision boundary, tool invocation, and memory transition must offer complete audit logs, replay capability, and automated regression testing.
Core Research & Engineering Thrusts
Swarm Coordination
Long-horizon planning, hierarchical task decomposition, multi-agent game-theoretic negotiations, and graph-structured state workflows.
AST-Native Code RAG
Deep syntactic parsing, call-graph topology indexing, and semantic flow modeling for precise monorepo code generation.
Empirical Verification
Automated red-teaming, zero-trust sandbox execution, and objective multi-turn task completion benchmarking.
Join the Community
We welcome researchers, system architects, and open-source contributors to collaborate on GitHub.