K.I.S.S(Keep It Static, Stupid): System prompt ft. cachingHow to build your system prompt to optimise for prompt caching & practical insightsApr 22, 2026·5 min read·96
Resources to transition to 'Applied AI' roleI have seen quite a few software devs(including me) with little to no background in ML jumping on the genAI train in the last 1-2 years. While this is a good thing and using LLMs with no prior classical ML knowledge helps in some ways it is also a de...Jun 27, 2025·3 min read·244
Curated resources: AI Product x How to eval?List of resouces for building a solid eval pipeline for your AI productMar 9, 2025·2 min read·525
Order of fields in Structured output can hurt LLMs outputEvals on how does the order of fields in prompts w/ structured output(JSON) affect the LLM response qualityJan 5, 2025·4 min read·3.3K
Experiments with gpt-4o vision and architecture diagramsEval based experimentation to figure out how well it works with architecture diagramsOct 22, 2024·4 min read·865
Notes on evals in LLM-based applicationsA big lesson I've learned while building LLM-based applications: "Unless you have a closed feedback loop, nothing you do will be good enough" Allow me to explain. When building an LLM app, I constantly found myself chasing 100s of different optimizat...Jul 22, 2024·2 min read·192
