15 years shipping production systems. Go, TypeScript, Kubernetes. Enterprise customer engineering at Apollo GraphQL. AI infrastructure at Atlas Health. Not a marketer who picked up a camera — an engineer who realized the most valuable developer content is the kind that can't be written by someone on the payroll.
Alderson.dev is an independent technical practice for developer tools companies: I bring independent technical judgment to Series A–C dev tools — building real implementations, publishing honest findings, and producing the third-party content internal teams can't credibly produce themselves.
Technical content packages built from real implementation work. The pattern is the same on every engagement: I use the product the way a senior engineer would, ship something that runs in production, and publish a written assessment that holds up to scrutiny.
The engineering and the writing are the same person. That's the whole pitch — a technical buyer can read the article, clone the repo, and trace every claim back to working code, and the person who wrote both is reachable on the same email thread.
I publish technical breakdowns through writing and my YouTube channel — that's where the independence is visible and proven. Client engagements are focused on engineering outcomes: working code, architecture clarity, and honest technical assessments that shape how your product is understood by the developer market.
Languages: Python (primary, AI/agent work), TypeScript (full-stack and Node services), Go (data infrastructure and high-throughput systems)
Infrastructure: Kubernetes, PostgreSQL, pgvector, Docker. Comfortable with managed platforms (Render, Modal, Vercel) and the underlying primitives both of them abstract over.
AI: Direct API integration to OpenAI, Anthropic, and open-source models. LangChain and the rest of the framework cycle when they earn their place — not by default. LiveKit for real-time voice agents. Playwright for browser-driven multi-agent workflows.
Production discipline: OpenTelemetry, structured logging, circuit breakers, retry logic, graceful degradation. Treating LLM calls like any other unreliable network dependency.
Sound architecture over framework churn. The interesting question is rarely which library — it's which parts of the problem actually need AI, and which parts are better solved by a SQL query or a state machine.
Based in Arizona. Truck in the desert. Operates nomadically. One engineer, not an agency — which means you work directly with the person producing the work, not a project manager farming it out.
I take on 2–3 engagements at a time. If you run a developer tools company with a strong product and thin third-party proof in the market, let's talk.
Book a Discovery Call →