TechHive Labs https://www.techhivelabs.net Lab notes on full-stack development, local AI, RAG pipelines, and the homelab experiments behind them. en-us Sat, 25 Jul 2026 00:00:00 GMT I Built a Bot Army to Attack My Own Game Server, and 45 of Them Took It Down https://www.techhivelabs.net/blog/game-server-scaling https://www.techhivelabs.net/blog/game-server-scaling Sat, 25 Jul 2026 00:00:00 GMT My movement bots said the server could hold hundreds. Then I taught them to swing a sword and the whole thing fell over at 45.. which is how I ended up shopping for CPUs instead of RAM. I Billed Myself for 6.7 Billion Tokens and the Total Came Out to $35.85 https://www.techhivelabs.net/blog/local-ai-true-cost https://www.techhivelabs.net/blog/local-ai-true-cost Tue, 07 Jul 2026 00:00:00 GMT My friends said the new model on my rig was drunk. Root-causing that from LiteLLM request logs turned into pricing every token my GPUs have ever produced.. and a $961 cloud comparison. I Run AI Agents Around the Clock and My Best Productivity Tool This Year Has No Battery https://www.techhivelabs.net/blog/de-digitizing https://www.techhivelabs.net/blog/de-digitizing Wed, 17 Jun 2026 00:00:00 GMT I built a homelab full of AI agents to run my life, then got actually organized with a pen, an A5 notebook, and a bullet journal key I bent to fit my markdown brain. I Chased Four Benchmark Points and Almost Missed the Model That's Twice as Fast https://www.techhivelabs.net/blog/local-ai-moe-vs-dense https://www.techhivelabs.net/blog/local-ai-moe-vs-dense Mon, 08 Jun 2026 00:00:00 GMT I finally got Qwen3.6 running locally, ran the dense flagship against the sparse MoE, and learned that four benchmark points are worth a lot less than 5x prefill. On an EPC Job, 'Sure, No Problem' Is the Most Expensive Sentence Onsite. An Agent Would Have Checked the Contract First. https://www.techhivelabs.net/blog/construction-agents https://www.techhivelabs.net/blog/construction-agents Tue, 02 Jun 2026 00:00:00 GMT A client asks your superintendent for a small favor. He says yes. Months later it's a claim nobody can reconstruct. The case for AI agents on EPC jobs. I Spent a Saturday Trying to Replace My Local AI to Save on Tokens. The Server Voted No. https://www.techhivelabs.net/blog/local-ai-upgrade https://www.techhivelabs.net/blog/local-ai-upgrade Tue, 28 Apr 2026 00:00:00 GMT Six hours, five models, one IPMI reboot, zero upgrades -- a field report on why the bleeding edge of local LLM hosting is mostly bleeding. I Spent $5K on GPUs Just to Learn That One GPU Was Enough https://www.techhivelabs.net/blog/local-ai-rig https://www.techhivelabs.net/blog/local-ai-rig Wed, 25 Mar 2026 00:00:00 GMT Four RTX 3090s, an EPYC server, and a 122B parameter model -- the journey from underwhelming to overkill and back to surprisingly elegant. I Spent 22 Years Programming Just to Fail at Making a Skeleton Swing a Sword https://www.techhivelabs.net/blog/thundoria-architecture https://www.techhivelabs.net/blog/thundoria-architecture Wed, 27 Aug 2025 00:00:00 GMT REST APIs to roguelikes: How I'm using a smaller game to learn the fundamentals before building my dream MMORPG. A honest indie gamedev journey. Build First - Learn Later https://www.techhivelabs.net/blog/vectly-scaling https://www.techhivelabs.net/blog/vectly-scaling Wed, 25 Jun 2025 00:00:00 GMT Building an AI app with zero knowledge of AI, hitting deployment walls, and evolving architecture to meet real-world needs. Building a Roguelike Game with Amazon Q https://www.techhivelabs.net/blog/q-roguelike https://www.techhivelabs.net/blog/q-roguelike Tue, 17 Jun 2025 00:00:00 GMT Using Amazon's AI coding assistant, Q, to build a browser-based roguelike with Nuxt.js — the potential and the limits of AI-assisted development.