Notes from running AI agents in production
What we learn deploying agents, automation and LLMs for universities, governments and enterprises — written by the people who build them.
LLMsChoosing an LLM for Enterprise Agents: The Model Is the Smallest Decision
Leaderboards change monthly. What decides whether an enterprise AI agent works is evaluation, data residency, cost per task and the freedom to switch models.
4 min read
Technology TrendsFive Enterprise AI Shifts to Plan for in 2027
Agents as colleagues, open tool standards, outcome pricing, governance by design and smaller private models: the shifts CIOs should budget for next year.
4 min read
AI AutomationAgents vs Automation: When Your Workflow Doesn't Need an LLM
Most of the automations we run in production have no language model in them at all. How to decide when a job needs an AI agent and when it needs a workflow.
4 min read
AI AgentsWhy Most AI Agent Pilots Never Reach Production
The demo works, the pilot impresses, and then nothing ships. Five reasons AI agent projects stall after the proof of concept, and what replaces each one.
5 min read
Want one of these agents running on your own metric?
Scope one agent against one number you already report on.
Or call +91 95525 86428