Anmol Bhardwaj
Writing about systems, product edges, and builder discipline.
Notes from shipping AI tooling, working through ambiguity, and trying to keep the engineering honest.
Featured5 min read
AI agents need operational constraints, not bigger prompts
The biggest reliability gains in agent systems usually come from tighter boundaries, not more clever instructions.
Mar 28, 20266 min read
Building Tailor-Talk from messy user intent
Natural-language scheduling gets hard the moment people stop speaking in clean forms and start speaking like people.
Mar 12, 20264 min read
Ship the interface before the infrastructure is perfect
You learn more from a sharp first interface than from two extra weeks of backend abstraction.