I design and build infrastructure systems with a focus on control, reliability, observability, and operational correctness.
My work centers on making complex, stateful systems — especially AI-driven ones — operable in real-world environments.

I care deeply about:

• Control planes and orchestration for complex, stateful systems
• Durable state, replayability, and auditability
• Operational safety: retries, fallbacks, policies, and human-in-the-loop interventions
• Clear system boundaries and explicit tradeoffs

Languages

• Python
• TypeScript / JavaScript
• Solidity
• Go (working proficiency)
• Rust (familiarity, exploring for systems)

Frameworks & Tooling

• Node.js, FastAPI, Express.js
• Ui: React, Nextjs
• Docker, Kubernetes, Terraform
• CI/CD: GitHub Actions, GitLab CI
• Structured logging, tracing, monitoring (Prometheus, Grafana, Sentry)
• Workflow orchestration: Prefect, Airflow
• Blockchain: Hardhat, Foundry

Domains & Expertise

• AI / LLM tooling and runtime orchestration
• Distributed execution & workflow systems
• On-chain → off-chain pipelines, Ethereum smart contracts

Tech stack includes some tools I am learning or experimenting with — demonstrating readiness for scalable, modern systems.

These projects are experimental — designed to explore system concerns rather than serve as full products.

• Prefer explicit state over implicit behavior
• Design for debuggability before scale
• Treat failures, retries, and human actions as first-class citizens
• Avoid hidden magic — systems should be explainable under pressure

Clarity always comes before cleverness.

• X / Twitter: @kelpejol
• Writing: Medium