AI Engineering.
Agent Systems.
Evolving Systems.
I build production-grade agent systems. Fast, measurable, and maintainable.
Context → orchestration → evals/guardrails/observability → infra.
ex-Squidrouter ($500M volume) · Led AI workflow adoption at FastLane Labs
For CTOs and eng leads rolling out AI-assisted dev without quality regressions.
For teams shipping agents that need evals, guardrails, and observability.
For protocol teams in audit or post-launch hardening.
Services
What you get in the first 2–4 weeks.
AI Engineering (Context, Evals, Workflow Design)
Turn ad-hoc prompting into a repeatable, evaluated engineering pipeline. Based on leading AI workflow adoption at FastLane Labs.
- Reusable context & evals pipeline from day one
- Claude Code / Codex CLI adoption across your team
- Cut PR review time while increasing defect catch rate
- Team trained and shipping with AI-assisted workflows
- Documented AI workflow architecture for your codebase
AI Agent Development
Ship an internal agent platform with orchestration, evals, and observability — not a demo, a production system.
- Production agent system with monitoring and guardrails
- Multi-agent orchestration with eval harness
- Agent-to-service integration with reliability guarantees
- Observability dashboards for agent behaviour and cost
- AI agent infrastructure for Web3 protocols
Protocol & Blockchain Engineering
Staff-level smart contract and backend engineering for Web3 teams in audit, post-audit, or launch phase.
- Audit-ready Solidity contracts (Foundry, Hardhat)
- Post-audit remediation and security hardening
- Go/TypeScript backend services for protocol infra
- Launch engineering from testnet to mainnet
- Cross-chain systems (EVM, Cosmos)
Track Record
Numbers first.
Co-founded Squidrouter (exited) — cross-chain swap protocol across 13+ EVM and 50+ Cosmos chains. View
FastLane Labs — plus shipped LST protocol core, audit remediation, and multi-region platform. View
Spearheaded Claude Code / Codex CLI adoption and agent-assisted workflows at FastLane Labs.
Building with Claude Code, Codex CLI, and agent frameworks since their earliest releases.
Selected Tools
Tools of the trade.
AI Engineering
Agent Systems
LLM Infrastructure
Protocol
Backend
Infrastructure
Process
How engagements work.
Discovery
30-minute callUnderstand your stack, goals, and constraints. Identify the highest-impact starting point.
Pilot
1–2 weeksHands-on sprint delivering a tangible outcome — a working agent, an eval pipeline, a hardened contract. You see the work before committing further.
Build & Rollout
2–4 weeksFull engagement with deliverables: repo templates, eval harness, monitoring dashboards, team training, and documentation.
Contact
Let's talk.