ServiceNow has released the entire platform documentation for AI systems. What is the use of it?

[Suggy]

ServiceNow has released the entire platform documentation for AI systems. What is the use of it?

this is what people say "One of the biggest hurdles when using AI to help develop on the ServiceNow platform is that AI often rely on outdated training data or "hallucinate" APIs that don't exist, meaning they will generate code that looks good, but doesn't actually work. They haven't had the full context of the latest platform releases until now."

then can anyone tell me how can I make use of that Markdown content to Ground answers using that like in chatgpt or copilot? so that I get authentic answers without any hallucinations?

Please answer only if you have 100% clarity. Thanks a lot in advance

[Suggy]

Anyone?

[Its_Azar]

Hello @Suggy 

Earlier, ChatGPT/Copilot etc answered from old training data, so they often hallucinated ServiceNow APIs or gave outdated syntax. truee

Now you can:

1. Store those Markdown docs in a vector DB
2. Retrieve relevant doc chunks for each question
3. Pass those chunks to the LLM as context
4. Force the model to answer only from retrieved docs

So instead of “answering from memory”, the AI answers using actual ServiceNow documentation. search retrieval etc

That’s the whole point of the release.

[Suggy]

Hi @Its_Azar Thanks for replying.

For point 1 and 2   -> Do you have the details of how that can be done?

[Its_Azar]

Hi there @Suggy 

Have you tried search retrieval as a tool for your agent, if not have a look.