Skip to content
View rlondner's full-sized avatar
26 followers · 4 following

Achievements

Achievement: Arctic Code Vault ContributorAchievement: Pull Sharkx2Achievement: Starstruck

Achievements

Achievement: Arctic Code Vault ContributorAchievement: Pull Sharkx2Achievement: Starstruck

Block or report rlondner

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
rlondner/README.md

Raphaël Londner

AI Strategy & Governance Leader — I turn ambiguous enterprise AI ambitions into production-ready, governed systems.

Master's in Engineering (École Centrale Paris) · 20+ years shipping technical products across the US and Europe · ex-MongoDB, Okta, Pure Storage · based in New York, NY.

I work at the seam between AI strategy and hands-on delivery: defining multi-year AI roadmaps, standing up RAG and LLM platforms in regulated industries (insurance, healthcare, legal), and building the governance, observability, and token economics that keep them running in production.

What I'm building & thinking about

  • Spec-driven LLM development — I built the same app twice, once with GitHub Spec-Kit and once with OpenSpec, to find out which workflow actually holds up. → doit_speckit-claude-opus · doit_openspec-core_claude-sonnet-4 · read the write-up
  • LLMs for regulated industries — fine-tuning and evaluating open models (Mistral-7B) for French-language insurance QA, with a reproducible eval harness. → ft-mistral7b-french-insurance-tests
  • AI applied to vertical data problems — smart reporting for commercial real estate. → cre-smart-report · why CRE brokers don't have a data problem
  • Parallel AI-assisted development — CodeSpawn, a Windows utility that launches multiple coding agents (Claude, Gemini, Codex, Qwen) side by side from a single File Explorer right-click, each in its own isolated workspace.
  • AI operations & observability — instrumenting agentic coding tools (OpenTelemetry, cost/latency/token telemetry) to make LLM usage measurable and governable.

Focus areas

Enterprise AI platforms · RAG systems · AI governance & operations · LLM observability & token optimization · Multi-tenant SaaS (insurance · healthcare · legal) · Spec-driven & vertical-slice development

Stack I reach for: TypeScript · Node.js / NestJS · Next.js / React · PostgreSQL · Python · LangChain · Azure OpenAI · Ollama · OpenTelemetry · Grafana

Writing

I publish on AI strategy, LLM engineering, and the changing shape of technical work.

Earlier developer-advocacy writing at MongoDB (serverless, Atlas, chatbots), Okta (identity/SAML), and Pure Storage (REST APIs, Kubernetes, CI/CD).

→ Full index on Medium

Connect

LinkedIn · Medium · X / @rlondner

Pinned Loading

  1. doit_speckit-claude-opus doit_speckit-claude-opus Public

    "DoIt" goal tracker built spec-first with GitHub Spec-Kit + Claude Code (Opus). Companion to my Spec-Kit vs. OpenSpec experiment. Next.js 16, React 19, TypeScript, PostgreSQL.

    TypeScript

  2. doit_openspec-core_claude-sonnet-4 doit_openspec-core_claude-sonnet-4 Public

    The same "DoIt" goal tracker, rebuilt with OpenSpec + Claude Sonnet 4 — the other half of the "I built the same app twice" experiment. Next.js 16, TypeScript, PostgreSQL.

    HTML 1

  3. CodeSpawn CodeSpawn Public

    Windows File Explorer utility to launch multiple AI coding agents (Claude, Gemini, Codex, Qwen) in parallel from any folder via right-click. PowerShell + Windows Terminal.

    PowerShell

  4. ft-mistral7b-french-insurance-tests ft-mistral7b-french-insurance-tests Public

    Reproducible benchmark comparing base vs. two fine-tuned Mistral-7B variants on French insurance Q&A. Local Ollama inference + Claude LLM-as-judge scoring across accuracy, completeness, French qual…

    Python

  5. cre-smart-report cre-smart-report Public

    Generates polished, image-rich commercial real estate reports from CoStar listing data — AI image classification (OpenAI CLIP) + python-docx layout automation.

    Python

  6. claude-code-otel claude-code-otel Public

    Forked from ColeMurray/claude-code-otel

    A comprehensive observability solution for monitoring Claude Code usage, performance, and costs.

    Makefile 1