Elisheba Builds·4d agoThe Model Proposes, Deterministic Code DisposesAsync buffers, strict boundaries, and where I keep LLMs out of the execution path
Elisheba Builds·Oct 27How I’d Let 5,000 Engineers Use LLMs Without Getting a Bank FinedBuilding a governed AI platform that helps engineers move faster without sacrificing security, risk management, or production reliability.
Elisheba Builds·Oct 6Metadata Is the Difference Between Retrieval and GuessingSimilarity search can’t tell an outdated guideline from the current one. Both look relevant. Metadata is what tells them apart.
Elisheba Builds·Sep 29My motto: Trace First, Eval Second: Building Evals for AI AgentsThe bug wasn’t the model. Some “AI eval” content stops at unit tests for prompts. Mine started somewhere messier: a fitness coaching agent…A response icon2A response icon2
Elisheba Builds·Sep 15How to cut AI costs with Model RoutingTL;DR: Default-routing every single user prompt to flagship frontier models is one of the fastest ways to burn your API credits. In this…
Elisheba Builds·May 28Engineering 100% Accuracy in a Real-Time AI Workout CoachLately, I’ve been working on Phase 1 of my optimization plan for Chaos Fit, a real-time AI workout coach built on top of Gemini Live and…
Elisheba Builds·May 14Engineers are tracking token costs and acceptance rates.The industry finally has a measurement conversation. Teams are watching token spend, AI code acceptance rates, and PR throughput. That’s…
Elisheba Builds·Apr 29How I Build: My Q2–2026 AI Engineering StackFor me, the shift from using AI to engineering with AI happened the moment I stopped reaching for tools and started building a system. Each…
Elisheba Builds·Apr 15My exact AI-assisted architecture review workflowI’m going to show you my workflow for exactly what I did on a real project, Global Impact Tracker, a tool I’m building with a Pro tier…
Elisheba Builds·Mar 25Reason #1,900 Why AI Won’t Replace Us: The Model Wasn’t Broken.I thought I broke my model.
