Personal AI Server

Avalon

Avalon is the nickname I have given to my in house AI server.

It has a Nvidia Tesla T4 GPU (which was data-center standard about 3 yrs ago)

I am running Ollama as the AI backend and have developed my own RAG system using SQL Server.
The source code for this is at https://github.com/benjaminemanuel13/Smile.Local.Ollama

I am using Phi-4 as my model of choice (it is small enough for home hosting but big enough to be useful).