Selected Deployments

Five deployments, one fixed format.

Deployment notes: field notes from taking AI to production. Each case study answers the same six questions in the same order: the problem and who owned it, the constraints, what I designed, the integrations, how it is governed and operated, and the outcome with how it was measured.

Every case study carries a responsibility tag so you can see what I built by hand, what I designed and shipped with AI agents under my review, and what I directed.

Qorinix inference control plane
Case study 1/5
Designed and shipped with AI agents under my review

A multi-tier model router across 12 LLM providers with streaming, a per-call cost ledger and usage-based billing, built for real-time agents, voice and trading alerts.

  • 12 LLM providers behind one API
  • Groq and Cerebras LPU inference for the low-latency tier
  • Streaming SSE
  • Per-call cost ledger and usage-based billing
LaSpend: read-only AI money assistant over UK Open Banking
Case study 2/5
Designed and shipped with AI agents under my review

Subscription detection, waste scoring and assisted cancellation over regulated Open Banking data, designed read-only by policy with human approval for every user-triggered action.

  • UK Open Banking API (read-only scopes)
  • Stripe subscriptions
  • Cloudflare Pages, Functions and D1
  • Multi-provider LLM routing with deterministic fallback
Fixxmi AI lead matching for a Swiss service marketplace
Case study 3/5
Designed and shipped with AI agents under my review

AI lead matching across 12+ service categories in DE-CH and EN, inside GDPR and nFADP boundaries, with pay-per-lead Stripe credit packs.

  • Next.js 15 front end
  • Firebase Cloud Functions
  • Firestore (europe-west6)
  • Stripe credit packs (pay per lead)
The governance loop applied to every AI call

Policy check, evidence-pack retrieval, prompt registry, model routing, action runtime, audit ledger, human-review escalation and policy feedback, wrapped around every AI call across the product portfolio.

  • Qorinix model router (12 providers)
  • Prompt registry with versioning
  • Content-addressed evidence packs
  • Audit ledger with per-call cost attribution
The philip.pm portfolio assistant (this site's chatbot)

A Next.js edge chatbot on Cloudflare Pages with a five-provider cascade, signed-cookie rate limiting with staged daily budgets and a curated preset answer bank streamed instantly. Live, inspectable proof of hands-on work.

  • Next.js 15 App Router on Cloudflare Pages (next-on-pages)
  • Edge runtime API route
  • Groq, Cerebras, Gemini, DeepSeek and OpenRouter
  • HMAC signed-cookie rate limiting

Want the numbers behind this one?

Adoption and precision figures are shared on a call, not published. Book 30 minutes or send a message.