AI Engineer building RAG systems, evaluation pipelines, and LLM-powered products.https://salimjordan.dev/https://salimjordan.dev/blog/building-a-socratic-ai-tutor-from-scratch/https://salimjordan.dev/blog/building-a-socratic-ai-tutor-from-scratch/Prompting a model to 'ask questions instead of answering' sounds like it should produce a Socratic tutor. Two recent preprints show why that assumption doesn't hold up, and what actually works instead.Fri, 03 Jul 2026 00:00:00 GMThttps://salimjordan.dev/blog/oatutor-and-the-pilot-study-that-reversed/https://salimjordan.dev/blog/oatutor-and-the-pilot-study-that-reversed/OATutor is an open-source adaptive tutor built around Bayesian Knowledge Tracing and built-in A/B testing. Two studies run on it, about fifteen months apart, show why that infrastructure matters. They also show how a small pilot's tidy result stopped holding up at scale.Fri, 03 Jul 2026 00:00:00 GMThttps://salimjordan.dev/blog/i-built-the-pipeline-the-math-worked-the-question-didnt/https://salimjordan.dev/blog/i-built-the-pipeline-the-math-worked-the-question-didnt/I tuned the formula constants until every spec gate passed. The pipeline still produced an empty high-priority bucket. The win wasn't a metric; it was recognizing the spec's two requirements were quietly antagonistic.Thu, 30 Apr 2026 00:00:00 GMThttps://salimjordan.dev/blog/refining-the-prompt-made-it-worse/https://salimjordan.dev/blog/refining-the-prompt-made-it-worse/I refined a sentiment classification prompt twice trying to clear an 85% accuracy target. Both attempts regressed. The win came from changing what I asked the model for, not how I asked for it.Thu, 30 Apr 2026 00:00:00 GMThttps://salimjordan.dev/blog/the-model-wasnt-random-it-was-backwards/https://salimjordan.dev/blog/the-model-wasnt-random-it-was-backwards/I expected a pre-trained embedding model to be neutral on dating compatibility. Instead, it was systematically wrong, scoring incompatible pairs higher than compatible ones. Here's why that made fine-tuning more interesting.Mon, 06 Apr 2026 00:00:00 GMThttps://salimjordan.dev/blog/reranking-isnt-always-better/https://salimjordan.dev/blog/reranking-isnt-always-better/I built two RAG pipelines. In one, adding a reranker dropped accuracy by 7%. In the other, it improved accuracy by 7%. Same technique, opposite results.Sat, 28 Mar 2026 00:00:00 GMT