/ newsletter · workshops · projects
Research, writing, and mentorship on agentic coding for engineers who ship.
I'm Sam Keen. I work hands-on with coding agents, then write and teach what actually holds up via weekly newsletter, live workshops, and the classroom. Author of Clean Architecture with Python, and formerly a generative AI architect at AWS.
/ newsletter
Every week, I dig into what's actually working in the new AI abstraction layer so you can make informed decisions without drowning in hype. A consistent weekly AI review for developers, plus long-form deep dives that go beyond the headlines.
Browse the archive on SubstackProduction-informed, not demo magic
Written by someone who shipped code for 25+ years
No hype, grounded in hands-on research and experimentation
A few things I've built alongside the writing. The newsletter shows how I think; these show what I ship.
An educational chat and RAG app for exploring retrieval patterns hands-on.
A teaching app that demonstrates a streaming chat interface over LLMs with retrieval-augmented generation. Flask backend, OpenRouter for model access, and a vanilla JavaScript front end. Cross-platform, and used as workshop material.
A notes app for research: plain Markdown in a folder you own, with an AI layer that finds the connections you haven't made yet.
My daily notes app, built to replace Obsidian. Beside every note is a ranked list of related notes it isn't linked to yet, and a model explains what they have in common. Keyword and semantic search, typed links (supports, contradicts), and answers grounded in your notes with citations you can check. Local first: embeddings run on your machine and chat uses Ollama by default.
A handheld, agent-assisted instrument for exploring the physical world.
Built on the M5Stack Tab5 (ESP32-P4): hold it, point it, move it, and ask questions while a person and an agent investigate something together in real time. ESP-IDF firmware on the device, a Mac-side Python investigation service, and local speech-to-text and spoken guidance. Co-developed with OpenAI Codex and Claude Code.
Teaching and showing up in the developer community: live, hands-on workshops, plus the talks, panels, and demos I give along the way.
always available
Learn the framework for structuring the project context file that makes Claude Code actually understand your codebase.
Watch free on MavenA live cohort workshop, run on a recurring basis. Build the context layer that turns Claude Code from a suggestion engine into a development partner: CLAUDE.md, skills, hooks, and the maturity ladder, all hands-on.
Join the waitlist on Mavenprevious
Packt · Deep Engineering · Jun 20, 2026
From prompts to systems. CLAUDE.md context layers, reusable skills, guardrails, and team-level practices for senior engineers and tech leads.
View event
In-person workshop
Hands-on workshop teaching retrieval-augmented generation patterns with real-world datasets.
Oregon AIR Challenge · Core Defender AI
Judged the pitch event for the inaugural Oregon AIR Challenge cohort, Core Defender AI’s statewide AI and robotics competition for students. A strong field of pitches made picking a winner genuinely difficult — looking forward to future cohorts.
Maseeh College of Engineering · Portland State University
Joined the AI Futures industry panel at the Maseeh Exchange: Events With Impact, alongside PSU faculty exploring how AI is being applied to real-world challenges.
Portland WIP AI Demos Lunch · UpStart Collective
Demoed the in-progress build at Portland’s informal show-and-tell for AI builders: a desktop inspector for every knob Claude Code gives you, where it lives, what it’s set to, and which layer wins.
/ about
I've spent 25+ years shipping production software at companies like Nike, Lululemon, AWS, and a handful of startups. I led the GenAI Innovation Lab at AWS, where I helped teams separate signal from noise in AI adoption.
Now my work is AI-assisted development: building in the open, writing, and teaching. Building with AI, not building AI. My newsletter reaches developers every week, and my workshops on Maven give teams hands-on systems for working with tools like Claude Code.
I also co-founded Portland AI Engineers, a community of 1,500+ practitioners exploring practical AI together.
I occasionally take on select consulting engagements for engineering teams navigating AI adoption.
Let's talk