Production agent harness
A Pi-based production harness with governed tools, persistent sessions, recovery, tracing, and model evaluation.
Read case studyCase studies
Selected work
Selected projects in AI agents, software, data platforms, and cloud infrastructure.
AI engineering
Engineering AI systems.
As an AI & Data Engineer, I build agents, execution tooling, and the data and cloud infrastructure that supports them.
Evaluation & loop engineering
Traces, feedback, evaluation, calibration, and replay evidence connected in a controlled improvement loop.
Read case studySmart Agent Factory
A custom durable graph runtime for human-gated, isolated, validated, and evidence-backed software delivery.
Read case studyAI platform & data foundations
Cloud infrastructure, data pipelines, observability, and access controls that make agent systems operable.
Read case studyAgent runtime engineering
Engineering a production agent harness with Pi
I evolved the Pi-based runtime and built major parts of its persistent execution, recovery, tracing, feedback, and evaluation systems.
Pi-based production harness
Read case studyDurable software delivery
Smart Agent Factory: durable delivery for AI-built software
I designed the system and implemented its custom graph, durable runtime, execution controls, human gates, validation path, and evidence-based delivery workflow.
Custom durable agent runtime
Read case studyMulti-agent orchestration
Building a verified multi-agent delivery platform
I designed and built the platform end to end, including the agent contracts, workflow state machine, verification gates, worktree isolation, persistence, API, and operational interface.
Artifact-gated agent delivery
Read case studyData platform engineering
Scaling a data platform on Kubernetes
I designed and implemented the core execution platform and operator, and co-designed or led the shared libraries and declarative orchestration tools.
Data platform modernization
Read case studyMachine learning infrastructure
Building an Airflow-driven ML platform on Kubernetes
I created the Airflow DAGs for data acquisition, model training, and model validation, and built the API layer that interacted with Airflow.
Acquisition to validation
Read case studyAnalytics cost optimization
Reducing monthly BigQuery cost by 25%
I implemented the optimization process across query behavior, physical table design, capacity, retention, and storage placement.
25% lower monthly BigQuery cost
Read case studyAI platform infrastructure
Building infrastructure and data foundations for AI
I built reusable infrastructure modules, agent scaffolding, and data workflows for AI services, including deployment configuration, access controls, observability, and backup automation.
Reusable infrastructure and data workflows
Read case studyAgent-assisted platform discoveryArchitecture proposal / prerequisites pending
Designing a conversational data-platform assistant
I proposed the conversational architecture, connected the platform's evidence sources to governed tools, and documented the prerequisites and open decisions for a pilot.
Read-only discovery architecture
Read case study