Resources
What we have learned, written down
Written by the people doing the work, mostly about the parts that turned out to be harder than expected.
Field notes
What we have learned running AI systems in production, written for engineers and the people who fund them.
2 articlesGuides
Longer playbooks and reference architectures. The frameworks we use in our own engagements.
1 guideAI glossary
Plain-English definitions with the trade-offs included. No hype, and no vendor framing.
11 termsLatest
Recent field notes
Write the evaluation set before you write the prompt
Most AI projects cannot tell whether they are finished, because nobody defined what correct looks like. The fix costs two days and changes everything downstream.
evaluationllmopspracticeThe DPDP erasure right will break your vector store
India's DPDP Act gives individuals a right to deletion. If your retrieval index has no lineage back to source records, you cannot honour it — and the fix is a full reindex.
dpdpgovernancearchitecture
Or skip the reading and just ask
Thirty minutes with an engineer beats an afternoon of research. Bring the process that annoys you most.
Or email [email protected] · we reply within 1 business day