Notes on AI agents that do real work.
Field notes on grounding, automation, support and sales, written while building AgentDeck.
The difference between a chatbot and an agent is what happens after the answer
Every vendor now says "AI agent". Most of them ship a better-worded chatbot. Here is the line that actually separates the two, and how to check which one you are buying.
Grounding, in practice. How to stop an AI agent from inventing answers
Hallucination is not a model problem you wait out. It is a retrieval and permissions problem you design around. A working checklist for grounding an agent in what your company actually knows.
Qualify leads in a conversation, not a form
Forms ask everyone the same questions in the same order and lose people at field four. A conversation asks the next useful question. Here is how to design one that still produces clean CRM records.
What to measure after you put an AI agent in front of customers
Conversation count is a vanity metric. These are the numbers that tell you whether the agent is helping, hiding problems, or quietly making things worse.
No articles in this topic yet.
See it working
on your own site.
Thirty minutes, your use case, a working agent on screen. No slides.
Book a 30‑min demo