---
title: "Discussion on \"How to build an AI sales assistant that does not make things up\" | Hashnode"
description: "Discussion on \"How to build an AI sales assistant that does not make things up\". Chatbot hallucinations are mostly a design problem. Four rules that make a business assistant answer from approved sources or say it does not know."
url: "https://hashnode.com/posts/how-to-build-an-ai-sales-assistant-that-does-not-make-things-up/6ab0d310ea331ee93db71587"
---

[Feed](https://hashnode.com/feed) /

Discussion

# [How to build an AI sales assistant that does not make things up](https://jarvisbitz.hashnode.dev/how-to-build-an-ai-sales-assistant-that-does-not-make-things-up?utm_source=hashnode&utm_medium=feed)

Chatbot hallucinations are mostly a design problem. Four rules that make a business assistant answer from approved sources or say it does not know.

Sep 21 · 10 min read · jarvisbitz.hashnode.dev  
Read the article

-   [#ai](https://hashnode.com/tag/ai)
-   [#llm](https://hashnode.com/tag/llm)
-   [#rag](https://hashnode.com/tag/rag)
-   [#architecture](https://hashnode.com/tag/architecture)

## Written by

[Suresh B](https://hashnode.com/@jarvisbitztech)

@jarvisbitztech · Ahmedabad, India

  
[View profile](https://hashnode.com/@jarvisbitztech)

Founder & CEO at JarvisBitz Tech

I run JarvisBitz Tech, an AI and software engineering company that builds production AI systems for clients, mostly in the US. More than a decade hands on across backend, cloud and SaaS. I write about what actually works when AI meets a real product: retrieval over large document sets, agents with human oversight, voice, and the engineering around the model that nobody puts in the demo.

## Responses 2

Markdown supported · ⌘↵ to post

[Mateo Ruiz](https://hashnode.com/@mateoruiz)

[Sep 21](https://hashnode.com/posts/how-to-build-an-ai-sales-assistant-that-does-not-make-things-up/6ab0d310ea331ee93db71587/comment/6ab0dc3bc706e0c5bccf740d)

The strongest point here is that `AGENTS.md` works best when it captures _repository-specific decisions_, not generic engineering advice.

We've seen the same pattern with AI coding tools: vague rules like "write clean code" barely change outputs, but concrete examples, forbidden patterns, and architectural constraints do. A short file with "Preferred/Avoid" examples often outperforms a massive document full of principles.

I also like the idea of the five-minute verification test. Removing the context file and comparing diffs is a practical way to measure whether your instructions are actually influencing agent behavior or just adding tokens.

One thing I'd add: include a small **"Why this exists"** section for non-obvious constraints (e.g., "don't simplify this retry logic because of production rate limits"). Agents are good at following patterns, but preserving past decisions is what prevents them from repeatedly proposing the same rejected ideas.

[Suresh B](https://hashnode.com/@jarvisbitztech)

[Sep 21](https://hashnode.com/posts/how-to-build-an-ai-sales-assistant-that-does-not-make-things-up/6ab0d310ea331ee93db71587/comment/6ab0dc3bc706e0c5bccf740d#reply-6ab0e24e6979badd2f5121dc)

Thanks for reading, [Mateo Ruiz](https://hashnode.com/@mateoruiz "Mateo Ruiz") . I think you may be thinking of a different article, this one is about business chatbots answering from approved sources rather than coding agents, but your point about concrete examples beating vague rules holds there too. A handful of real example phrasings per intent does more for classification than any amount of instruction text.
