Inspiration
What it does
How we built it
The idea would be to export data from crunchbase about pre seed startup fundings, and then extract features from these startups to create a machine learning model that links the post seed valuation of a startup to these features. The multi agent part would be to use LLMs to extract features from the startup description for example
Challenges we ran into
Getting data from crunchbase wasn't straightforward, there's not much post seed valuation information, scraping with multi on and llamaindex didn't work
Accomplishments that we're proud of
Getting to test all the new products from the hackathon partners
What we learned
Make sure you'll get the data before starting a project
What's next for Venture Pulse
Getting some data
Log in or sign up for Devpost to join the conversation.