[FR]: Add Support for Additional AI Provider Configurations

[madhankumar2211]

Proposal summary

Currently, the platform supports integrating AI providers like OpenAI in the configuration page for the playground. To improve flexibility and cater to a broader user base, we propose adding configurations for additional AI providers, specifically for locally hosted models such as those managed by Ollama and LM Studio.

Motivation

1. Offer flexibility in choosing between cloud-based and locally hosted solutions.
2. The existing setup limits users to cloud-based AI providers, which can be restrictive due to internet dependency, high costs, and potential data privacy concerns.
3. Users with offline environments or sensitive workflows need an option to leverage locally hosted AI models seamlessly.

[jverre]

Hi @madhankumar2211
This is a great idea !

@aadereiko Can we look into this once we have finished the dataset integration for the playground - The idea would be to have an AI provider that doesn't go through the BE proxy but rather Ollama running locally (see https://github.com/ollama/ollama?tab=readme-ov-file#chat-with-a-model)

[aadereiko]

Hey @madhankumar2211 @jverre

A clever remark. Sure, I'll get to it right after I finish dataset integration :)