I build on the web, but I'm increasingly drawn to the why behind what we build. Moving from full-stack engineering toward AI Product Management — where user problems, model capabilities, and shipping velocity all collide.
- 🔭 Currently learning Claude Code in public — one chapter a day
- ✍️ Writing short reflections on AI products and the engineer → PM shift
- 🧠 Exploring how LLM-native UX is changing the PM playbook
- 🎯 Looking for Associate PM / AI PM roles where engineers-turned-PMs are welcome
| Area | What it looks like in practice |
|---|---|
| AI product sense | Reverse-engineering features in Claude, Cursor, Perplexity and asking why that tradeoff |
| Shipping | Small, weekly experiments — each with a hypothesis and a written outcome |
| Writing | Learning logs, product notes, and "what I'd change" takes |
| Engineering depth | Still ship full-stack side projects so I stay close to how things actually get built |
- 📚 Claude Code, day by day — progress log + daily posts on LinkedIn and X
- 🔬 AI product notes — one product, one flow, one tradeoff per post
- 🛠️ Side projects — React / Node apps where AI is the feature, not the gimmick
Product: Notion · Figma · Amplitude (learning) · Linear AI tooling: Claude Code · ChatGPT · Cursor · prompt design Engineering: JavaScript · React · Node · Express · MongoDB · Firebase
- How LLM-native products break the traditional PRD → ship → measure loop
- Why context and prompts are the new UX — and how PMs should design them intentionally
- How engineer-PMs can use tools like Claude Code to prototype before speccing
- The gap between demo-magic and production-reliability in AI features — and who owns closing it
Engineer who writes. PM in training. Always down to talk about AI products.
📬 iamaman526@gmail.com