Everything I ship. Memory infrastructure, companion robots, commerce — all production, all documented. The short videos point here. The guides explain how.
Migrated from the original aldowen.com landing — every link preserved, same URLs.
Active · ProductionSemantic memory server for AI agents. 19,500+ facts indexed, 70–98% token reduction, FTS5 instant recall — all local, no GPU, SQLite + bge-m3. The memory layer this site runs on.
Automated memory management with tiered retention and Obsidian sync. Raw interaction logs → curated wisdom.
Semantic triples, vector search, procedural memory. Next-gen agent memory — all free, all local.
PKM via the ACE framework and AI-generated Maps of Content. Obsidian-native synthesis.
One-command installer for a production AI gateway. Your brain, always online, Tailscale-hardened.
Practitioner's guide to high-availability gateways and agent infrastructure hardening.
Keeping agents alive through deploys, restarts, and network partitions.
Honest reviews of 2026's best AI coding tools — Cursor, OpenCode, and the rest.
4WD companion robot. BLE + CV + LLM. The original — predecessor to OttoBot.
ESP32-S3 with browser-ROS bridge, OV5640, audio streaming, MCP server (10 tools), 5-axis mood engine.
Next iteration — on-device inference, tighter ROS loop.
WhatsApp commerce — dashboard, OTP login, Midtrans payments, live in production.
Lightweight agent runner — quick deploys, low overhead.
Who I am, how I build, and why agents should remember.
/blog/atom.xml