OpenAI - API Assistants

[sullen totem]

Hi Devs

I've recently built an AI assistant using OpenAI to handle FAQs & Information for my product. I don't have much experience with Python, and I'm a bit confused about using the dashboard, especially when it comes to completions or batches and threads inside the assistant

I'm not sure if I need to do all the coding or set up the assistant in a specific environment, or if I can just handle everything through the dashboard and then integrate it into my code.

I've also been trying to make the bot generate new response messages instead of just copying and pasting static content. I want it to use data to interact with users intelligently, but I haven't had much luck so far. I've tried searching for solutions but haven't found anything helpful.

When I tried fine-tuning the data, the results were poor. The bot was giving incorrect responses that weren't even related to the information I provided.

If anyone could help me understand how the dashboard and coding work together, or recommend a helpful YouTube video, I'd really appreciate it. I'm just looking for some guidance on where to start on OpenAI API

[little fable]

dashboard chat/assistants are mainly just for prototyping and testing

setting one on your own server requires some coding and security knowledge at minimum

[sullen totem]

Thanks, but still have a few concerns, all features available is just for testing? such fine tuning? or storage? file search?

Do you have any videos/reference to start working on it and learn about it more?

[little fable]

dashboard has many features but not all of the things that the api offers and you cant use all the options there

https://platform.openai.com/docs/overview

[sullen totem]

Which is useful and which is not?

[little fable]

useful for what exactly?

[sullen totem]

For my case, of creating the chatbot for FAQ & Product Information Tutor, which can save time instead of code it or implement it in the code.

[little fable]

you can setup and test the assistant there but the actual use always happens via api calls that you need to code and add to your site

then you also have to comply to rate limits, handle safety so the users dont ask how to make a bomb or create massive costs in api fees, etc.

[sullen totem]

I mean for example, I cannot add anything to fine tuning, it must be in my code? even the storge files?

[little fable]

you can fine-tune there

[tame sail]

I use the Assistant back end to define my bot, instead of defining it in the the code itself. Descriptive answers in your system prompt will do wonders.

[sullen totem]

Really? What kind of things you manage through API Assistant OpenAI instead of the code itself? do you keep integrate the functions of the assisntant insde the bot? i am actually aiming to have chatbot serve as product information and faq bot, to add in my mobile app for users to get help

[tame sail]

You can. Hit me up in dm I can tell you all about it. I love it.