My Portfolio https://davidokpare.com This is my portfolio RSS feed RL 101: Building An Environment https://davidokpare.com/blog/rl-101-q-learning A top-down approach to building intuition in reinforcement learning starting with building an environment. Mon, 15 Dec 2025 00:00:00 GMT Building a Research Agent https://davidokpare.com/blog/building-a-research-agent Building a multi-agent research system using the orchestrator-worker pattern, where subagents work as intelligent tool calls to handle complex research queries. Sun, 24 Aug 2025 00:00:00 GMT Not One Prompt Rules Them All https://davidokpare.com/blog/not-one-prompt-rules-them-all Why your fallback models need fallback prompts and how to preserve output quality when models switch Tue, 19 Aug 2025 00:00:00 GMT Bridging the Three Gulfs of LLM Development: A Primer to Evals https://davidokpare.com/blog/a-primer-to-evals How to bridge the three gulfs of LLM evaluation (comprehension, specification, and generalization) to build reliable AI systems that actually work in production. Sun, 27 Jul 2025 00:00:00 GMT Building a Multimodal Search Engine: A Reddit Case Study https://davidokpare.com/blog/building-multimodal-search-engine-reddit-case-study Exploring multimodality to enhance semantic search capabilities with image and text Thu, 07 Nov 2024 00:00:00 GMT Building a Vector Database on SQLite, Numpy and KNNs https://davidokpare.com/blog/building-a-vector-database-on-sqlite-numpy-and-knns Minimalist implementation of a vector database on SQLite, Numpy and KNNs. Thu, 10 Aug 2023 00:00:00 GMT