Add Ollama LLM Provider

[jspada200]

Just like Gemini and OpenAI. Add an Ollama provider to access a model running locally

[Aneesh-0108]

Hello there @jspada200 ,I would like to work on this issue related to adding ollama provider,can u please assign this issue to me?

Thank you!

[jspada200]

Thank you for volenteering @Aneesh-0108 ! Please review https://github.com/AcademySoftwareFoundation/dna/wiki/Your-First-Contribution

Before you submit your PR, please test and ensure it builds and works properly.

[a-chr1s]

Hi @jspada200

Happy to take this on if it's free. I have a solution working in the UI and am just working on the pytest file. Can push a PR once that has been tested. Hopefully should be in the next few days.

Kind regards,

Chris

[jspada200]

Thanks @a-chr1s ! Let me know if you have any challenges with setting up the development env. I tried to make it easy but some have run into issues. Would like to make it as approchable as we can. Thanks!

[jspada200]

Just another note @a-chr1s , this PR abstracts some of the openai provider out which directly affects this work. I would wait till this is in. Going to work on getting it in ASAP: #132

[a-chr1s]

Thanks @jspada200 I'll take a look and see if I can pre-empt some of the incoming changes.

Re: Development environment set-up, the only issue I faced was with Vexa doing transcription on the mock SG database setup. I can open a GH issue but the jist is that Vexa was getting a 403 with the Environment Variables set in the Dockerfile. It appears that it may require other ENV vars set? Tested this on M5 MBP macOS 26.3.1 as well as Rocky 9.7 (VM). This could just be something that needs a documentation change

[RoseAllenM]

Gemini PR has been merged.

I was having trouble with Vexa as well. Instead of investigating that, I created this PR which adds a script to pre-seed a transcription using the sample data @camerontarget14 already added to the repo.
#143