Aug 08, 2025, 01:15 PM
A few months ago, I started developing a rather curious project! While using those APIs that query personal information, emails, etc., a thought crossed my mind... What if this query interaction was handled by a conversational LLM agent?
With that in mind, I began developing an application designed to run locally (no OpenAI API LOL) with a single-core AI model (Gemma 3) because my PC isn’t the most powerful... I’ve already finished most of the logical configurations (aka backend), integrated the model with contextualized information from leaked databases via RAG, and got some pretty cool results! My next steps are to integrate this with deep web search techniques to create a much more detailed experience for the results!
What do you think of the idea? At the very least, it’s a "fun" project to work on, lol.
If anyone wants to contribute by sharing databases rich in information, I’m open to it! It’s quite laborious to gather the necessary pieces to access some of these databases... And even worse, some are missing data, and I don’t know where to find it!
With that in mind, I began developing an application designed to run locally (no OpenAI API LOL) with a single-core AI model (Gemma 3) because my PC isn’t the most powerful... I’ve already finished most of the logical configurations (aka backend), integrated the model with contextualized information from leaked databases via RAG, and got some pretty cool results! My next steps are to integrate this with deep web search techniques to create a much more detailed experience for the results!
What do you think of the idea? At the very least, it’s a "fun" project to work on, lol.
If anyone wants to contribute by sharing databases rich in information, I’m open to it! It’s quite laborious to gather the necessary pieces to access some of these databases... And even worse, some are missing data, and I don’t know where to find it!
![[Image: 128.gif]](https://i.ibb.co/hJ2HmDxK/128.gif)