Consulting services

Independent reviews from someone who builds and measures this work hands-on. Every engagement has a written scope agreed before it starts and ends with a prioritised list of findings.

Scoped reviews

AI Gateway Readiness Review

For teams putting model and agent traffic through a gateway, or about to.

  • Map current model and MCP tool traffic and which gateway or path carries each
  • Check where policy binds and who can change it
  • Test how an agent is identified, registered, and stopped
  • Written findings with a recommended operating model
Ask about this review

RAG Security Assessment

For a retrieval-augmented assistant that is in production or close to it.

  • Walk the ten areas of the RAG security checklist against your pipeline
  • Check retrieval-time access control and injection exposure
  • Review PII handling, audit logging, and the eval set
  • Written findings ranked by risk, with fixes
Ask about this assessment

Other areas

AI gateway and agent governance

How model calls and agent tool calls are governed: which gateway carries which traffic, where policy binds, how the registry works, and how fast one agent can be stopped.

Related reading

RAG security review

Access control at retrieval, prompt-injection exposure, PII in the index, audit logging, and evals, checked against your pipeline rather than a generic list.

Related reading

Agent evaluation and observability

Which parts of an eval framework to build, which to buy from a managed service, and how evals, tracing, and alerting fit together in production.

Related reading

Data architecture review

An independent look at pipelines, storage, access patterns, and cost on Google Cloud, and how well the platform positions you for AI workloads. Recommendations you can act on, not a vendor pitch.

Related reading

Governance and metadata

Data catalogs, ownership and stewardship, lineage, and data products, set up so people keep using them.

Related reading

Team and career coaching

One-on-one sessions for data engineers and architects working through the shift to AI: portfolio and GitHub feedback, interview preparation, and honest career advice.

Related reading

How it works

  1. Send a short description of what you're working on through the contact form.
  2. I reply within one to two business days to discuss scope and availability.
  3. We agree a written scope and format: a one-time review, a short series of sessions, or ad-hoc hours.
Start a conversation