I design and build autonomous systems that use software, and structured workflows to perform real technical work.

• Autonomous issue discovery
• Codebase investigation and localization
• Patch generation
• Regression-test generation
• Automated testing and evaluation
• Repository-specific policy enforcement
• Failure detection and recovery
• Integrity verification
• Evidence and audit trails
• Measurement of autonomous engineering workflows
• Controlled autonomous execution

• Agent orchestration and tool use
• Multi-stage reasoning and execution
• Persistent state
• Decision-making systems
• Policy enforcement
• Failure detection and recovery
• Budget and concurrency controls
• Exception handling
• Human/AI collaboration
• Systems designed for continuous operation

• Mechanism design
• Token economics
• On-chain coordination
• Protocol architecture
• Crypto-native infrastructure
• Capital-flow intelligence
• Autonomous systems for crypto

Omega is my primary engineering project: an autonomous software engineering system that I designed and built from scratch.

Omega combines systems engineering, software tooling, orchestration, evaluation, policy enforcement, measurement, persistent state, and controlled execution into an engineered software-development workflow:

Discover → Investigate → Implement → Test → Evaluate → Verify → Review → Ship

Execution authority is determined by repository policy, machine-verifiable requirements, system integrity, and configured autonomy boundaries.

The surrounding architecture is the engineering work.

That includes:

• Issue discovery and triage
• Codebase localization
• Patch generation
• Regression-test generation
• Independent patch evaluation
• Repository-specific contribution policies
• Automated verification
• Blinded engineering estimates
• Outcome measurement
• Append-only audit trails
• Cost attribution
• Failure recovery
• Integrity verification
• Budget and concurrency enforcement
• Exception handling
• Persistent operational state

The objective isn't simply to generate code.

It's to build an autonomous system capable of performing useful engineering work while remaining:

measurable · constrained · auditable · recoverable

I'm particularly interested in:

• Autonomous software engineering
• Agent evaluation
• Systems engineering
• Human/AI collaboration
• Reliability and failure recovery
• AI governance and safety boundaries
• Economic measurement of AI labor
• Autonomous infrastructure
• Agent orchestration
• Systems that turn AI capabilities into repeatable workflows

I also build and research crypto-native systems around:

• Token and incentive design
• Mechanism engineering
• On-chain coordination
• Permissioning and governance
• Protocol upgrades
• Secure execution
• Capital-flow intelligence
• Autonomous crypto infrastructure

My broader interest is where incentives, coordination, software, and autonomous systems intersect.