Infrastructure and platform engineer with production experience building internal developer platforms, automation systems, and agentic AI tooling. Currently at SITA (Reading) working on cloud infrastructure and developer productivity. MSc Distinction in Advanced Computer Science. Research background in agentic LLM self-correction, ML observability, and verification-guided reasoning systems.
Core stack: Go · TypeScript · Python · Kubernetes · Terraform · Azure · GCP
Building Krynix — an open-source agent behavioural trust layer for agentic systems.
Agent behavioural trust verification toolkit. Traces agent action sequences, detects anomalies in decision trajectories, and acts as a CI gate to block deployment of agents exhibiting unexpected behaviour. Framework-agnostic, local-first, zero data exfiltration.
Python · TypeScript · CLI · CI/CD integration
Verification-guided LLM reasoning system. Couples a generator with typed verifiers — arithmetic, symbolic algebra via SymPy, formal logic — and an iterative controller that runs detect → verify → revise loops. Supports OpenAI, Anthropic, and local inference via Ollama.
Related to MSc dissertation research in agentic self-correction.
Python · FastAPI · Docker · Kubernetes
MLOps observability platform for ML event ingestion, drift and fairness monitoring, and alerting. ClickHouse-backed with Python and Node SDKs, CLI tooling, RBAC, and SSO via OIDC.
Python · FastAPI · ClickHouse · Prometheus · Docker
Languages: Go · TypeScript · Python · JavaScript
Infrastructure: Kubernetes · Terraform · Docker · Azure · GCP · AWS · Linux
Backend: REST APIs · gRPC · Microservices · Event-driven architecture · Distributed systems
Frontend: React · TypeScript · HTML · CSS
CI/CD: GitHub Actions · Azure Pipelines · Testing automation · Release pipelines
Observability: Prometheus · Metrics · Logging · Distributed tracing · Incident response
Databases: PostgreSQL · MySQL · ClickHouse · MongoDB
AI and Agentic Systems: LLM integration · MCP · Agent workflow design · Agentic self-correction (MSc research) · Ollama
ML and Data: PyTorch · scikit-learn · Pandas · NumPy · MLOps · Drift detection · Fairness monitoring · ClickHouse-backed event pipelines
Open to interesting problems in infrastructure, developer tooling, and agentic systems.