← sources

Agency and bot builder question

source: reddit (reddit:AI_Agents)published 2026-10-03 06:28ingested 2026-10-04 02:35

retained text

I'm a co-founder building testing tools for conversational AI agents. Most teams I've talked to still test their bots by hand, and they redo the whole round every time a prompt or the knowledge base changes.

I'd like to understand how this works for agencies juggling several clients:

1. How do you test a bot before handing it over?  
2. When a client asks for changes after go-live, do you retest everything or only what changed?  
3. Is testing a line item in your quotes?

You can answer in the comments, and I'd also love to do a few 20-minute research calls with agency owners and builders. I won't pitch anything, and I'll share a summary of what I learn with everyone who takes part.

Comment or DM me if you're up for a chat. Thanks!

document fetched 2026-10-04 00:19 · canonical https://reddit.com/r/AI_Agents/comments/1wwg4sr/agency_and_bot_builder_question