Vector search API without using a library

Hey guys,

I tried to get an answer from the support,
but they didn’t understand me,
Hopefully you will.

I am building a chatbot that I want to connect to AI using a knowledge base
and use MongoDB as the Vector DB.
and use the vector search by API.

To do it, I need API access to query the vectorDB.
And I can’t install libraries in the Chatbot platform
only solid js code, that’s it.
so an API request will be a fetch(endpoint, params) type of thing.

Everything in your docs works with an installed library,
Without the library, I need to know the API structure to create a fetch request in the chatbot platform.
I hope that helps to understand,

Thanks for the help in advance.

[quote=“Vitrue_Your_Innovation_Our_Art, post:1, topic:288574, full:true”]HPInstantInk
Hey guys,

I tried to get an answer from the support,
but they didn’t understand me,
Hopefully you will.

I am building a chatbot that I want to connect to AI using a knowledge base
and use MongoDB as the Vector DB.
and use the vector search by API.

To do it, I need API access to query the vectorDB.
And I can’t install libraries in the Chatbot platform
only solid js code, that’s it.
so an API request will be a fetch(endpoint, params) type of thing.

Everything in your docs works with an installed library,
Without the library, I need to know the API structure to create a fetch request in the chatbot platform.
I hope that helps to understand,

Thanks for the help in advance.
[/quote]

Hello,
I’m also facing an issue with building a chatbot that connects to AI using a knowledge base and MongoDB as the Vector DB. I need API access to query the Vector DB with fetch requests since I can’t install libraries on the chatbot platform. Your docs mostly require installed libraries, so I need the API structure to make this work. Thanks for your help!

Best regards,
florence023

Hi @florence023!

I am building a chatbot that I want to connect to AI using a knowledge base

What do you mean by knowledge base here? are you referring to AWS Bedrock knowledge base? This allows you to connect your data in s3 to Atlas vector search via AWS Bedrock, and you get an API endpoint that can be used to query directly

If you are trying to build a custom chatbot in JS , have you tried the LangChain JS? Here’s the [tutorial for building a chatbot in JS|https://mongodb.prakticum-team.ru/docs/atlas/atlas-vector-search/ai-integrations/langchain-js/]