PrometheusAI

• 

  PrometheusAI (UI)

• 

  PrometheusAI (Logo)

• 

  Available on GitHub

Inspiration

Privacy, Token limits on closed LLMs, APIs

What it does

PrometheusAI works completely offline on your system, giving you option to run a llm at your local server rather than spending on APIs

How we built it

We used Python, FlaskApp, LM Studio, Frontend (HTML/CSS/JS), Google Cloud (Auth)

Challenges we ran into

Anonyms chat option, which keeps the conversation in the RAM and once the user ends the chat/conversation, it discards it from the RAM.

Accomplishments that we're proud of

We built multiple automations using PrometheusAI setup (Linkedin Job Apply, Linkedin Comment Agent, YCombination Pitch Agent)

What we learned

You don't need to spend millions to host a Local LLM at your server

What's next for PrometheusAI

Deploy it with RAG (Retrieval Augmented Generation) and make LLM aware about our clients company data to make it full fletched local llm for our clients

Note: the given link will only work when I am keeping my server live at my end

Hosted at: https://github.com/SamirSengupta/PrometheusAI

Built With

• deep-learning
• flaskapp
• lm-studio
• machine-learning
• neural-networks
• python