Figuring out a better way to search through texts and understand who, what, and how we got to that point in a conversation is better for understanding relationships and possibly realizing why you did certain actions in the past.

What it does

Currently, it semantically searches and retrieves context from IMsg conversations given a very vague prompt about the things you're interested in (i.e. tell me about a time that I was sad with a given person"). Over time, the hope is to make this more specific and allow for queries into why you may need to go to a certain event for example.

How we built it

Python + SQLite + Sentence Transformers + Twilio

Challenges we ran into

Parsing Imsg is pretty hard, but scoping is difficult.

Accomplishments that we're proud of


What we learned


What's next for IMessage Search and Relationship Management

Implementing LSM and Sentiment Analysis to greater detail. Recursive queries using an LLM to generate better search responses that are more suited for the embeddings.

Built With

Share this project: