Data engineer based in the UK. I build small, free, on-prem tools for data and SQL work, and send fixes upstream to the projects I use. MSc Data Analytics, Aston.
Portfolio · Governed Agent Stack · LinkedIn · PyPI
- governed-agent-stack: free, on-prem building blocks for an AI agent you can point at a real database and audit. schema-scout maps the data, sql-explorer-mcp and sql-sop fence in read-only access, FloorMind reasons, and agent-blackbox keeps a tamper-evident record.
- FloorMind: on-prem natural-language-to-SQL over manufacturing data. A LangGraph agent, a local LLM via Ollama, sql-sop validation, and a tamper-evident audit log. Eval-measured.
- agent-blackbox: an append-only, hash-chained ledger that gives agent actions a tamper-evident audit trail. On-prem, zero dependencies.
- sql-explorer-mcp:
read-only MCP server for SQL Server, Postgres, and SQLite, with a layered
safety stack.
pip install sql-explorer-mcp - query-warden: role-based access control for SQL. Decides whether a role may touch the tables and columns a query references, before it runs. On-prem.
- schema-scout: reverse-engineers a SQL Server database into an AI-ready catalog. Infers the undeclared foreign keys, flags PII, scores how AI-ready the schema is, and renders an offline dashboard. Runs entirely on-prem.
- sql-sop: a SQL linter
that catches dangerous and non-conforming SQL in CI and pre-commit.
pip install sql-sop· browser playground - sql-sop-mcp: an MCP
server wrapping the linter, so Claude / Cursor / any MCP client can call it.
pip install sql-sop-mcp - pr-sop: small PR-governance
checks: changelog drift, version mismatch, stale pre-commit pins.
pip install pr-sop - morning-brief:
rule-based Gmail triage, read-only, no LLM.
pip install morning-brief
- production-analytics-pipeline: incremental ETL from a manufacturing ERP. FastAPI + Next.js + Power BI, Prefect orchestration, 53 tests.
- uk-crime-pipeline: Police UK API into PostgreSQL and BigQuery. dbt marts, ~100k records, Polars ingestion.
Python · SQL / T-SQL · dbt · SQL Server / PostgreSQL / BigQuery · pandas / Polars · FastAPI · Ollama / LangGraph (on-prem AI) · MCP · Docker · GitHub Actions
Learning (2026): Microsoft Fabric · KQL · Azure · Snowflake