Stop Fine-Tuning to Fix Retrieval Problems — Anant Srivastava
Enterprise teams default to fine-tuning by accident, not conscious architectural choice
“most corporate teams make this decision by accident”
Anant Srivastava argues that the core architectural decision in enterprise AI is how knowledge is brought to inference — via prompt context, retrieval/memory, or model fine-tuning — but most teams make this choice accidentally by escalating incrementally when answers are wrong. He frames these as three distinct tools for distinct tasks, not a ladder to climb, and uses a real internal support assistant example to show how ad-hoc decisions accumulate into an unplanned architecture.