https://anasarfeen.dev/blog Notes on building reinforcement-learning robotics, LLM-driven agents, and full-stack platforms. en https://anasarfeen.dev/blog/hybrid-brain-for-competitive-pokemon https://anasarfeen.dev/blog/hybrid-brain-for-competitive-pokemon Sat, 25 Jul 2026 00:00:00 GMT Pure heuristics miss deep strategy. Pure LLMs are too slow for every turn. Here's why poke-ai splits the decision loop between the two instead of picking one. https://anasarfeen.dev/blog/building-amazecc-replacing-vtop https://anasarfeen.dev/blog/building-amazecc-replacing-vtop Sun, 05 Jul 2026 00:00:00 GMT Notes from building a ~130-route production platform that real VIT students use instead of VTOP — on why the hard part was never the dashboard. https://anasarfeen.dev/blog/curriculum-learning-warehouse-rover https://anasarfeen.dev/blog/curriculum-learning-warehouse-rover Thu, 02 Apr 2026 00:00:00 GMT Why a warehouse rover trained directly on hard layouts never learns anything, and how promoting it through four difficulty tiers at an 80% success threshold fixed that.