Four ways to bring me in.

Architecture Review · 1 week

Async work + 2 calls. You get a written report covering: findings, prioritised fix plan, cost-ceiling guidance, and a 30/60/90-day roadmap. Most clients use this to decide what to do next, with or without me.

You'll get

• Architecture diagram (current state)
• Top 5 risks ranked by impact and effort
• 30/60/90-day fix plan with rough effort estimates
• Cost-ceiling and scale guidance
• Recommendations for what to build, what to buy, what to drop

Not included

Hands-on coding. (For that, you want Tier 3.)

Book a Review

Cost & Scale Audit · 2 weeks

Everything in Tier 1, plus a full AWS-bill breakdown by service, a quick-win script you can run today to recover obvious savings, and a phased scale roadmap with effort estimates.

You'll get

• Architecture review (Tier 1)
• AWS bill broken down by service, environment, and team
• Cost-per-user and cost-per-request metrics
• Quick-win script: idle resources, unattached volumes, oversized instances, missing lifecycle rules
• Scale roadmap: where to cache, where to queue, where to horizontally scale

Discuss the audit

Modernization Sprint · 6–12 weeks

Embedded part-time. I refactor or migrate one critical module end-to-end and leave a playbook the rest of the team uses to do the same to the rest of the codebase. Hands-on coding throughout.

Examples I've delivered

• 40-microservice mess collapsed into a single FastAPI modular monolith (Inara — 6 weeks)
• Cookie-bloat → JWT + server-side RBAC migration (Jobscope — 4 weeks)
• Spring Boot → Python FastAPI port using AI-assisted translation (Crimson — 6 weeks)

You'll get

• One critical module completely modernized in production
• A repeatable playbook for the rest of the codebase
• Architecture decision records (ADRs) for every non-obvious choice
• Trained team — every PR I write becomes a learning resource

Talk about a sprint

Fractional Solution Architect · 3+ months

1–2 days/week, ongoing. Architecture decisions, design reviews, hiring help, and direct hands-on work on the most critical pieces. Best fit for fast-growing teams that need senior architecture judgement but can't (yet) hire it full-time.

Typical activities

• Weekly architecture review meeting
• Design docs / RFC reviews on every major change
• Hands-on code on critical pieces (auth, data model, payment, scale-critical paths)
• Hiring panel for senior engineering roles
• Cloud bill review every quarter

Discuss fractional

AI Pilot · 4–8 weeks

One AI feature shipped to production with rule-based fallbacks, an evaluation harness, and cost guardrails. For organisations that want AI in their product but have been burned by demos that don't survive contact with real users.

You'll get

• One AI feature live with real users (e.g. lead qualification, support deflection, document summary, code analysis)
• LLM-agnostic provider layer with fallback (OpenAI / Anthropic / Gemini)
• Deterministic rule layer for high-stakes decisions
• Evaluation harness — you can A/B prompts and measure regression
• Cost ceiling per request and per user

Talk about an AI pilot