Build your AI Startup Hackathon. Episode 3

We are happy to welcome you on AI Hackathon on lablab.ai! Let's create and build together over next week!

Hackathon Kick-Off

We will talk about the Hackathon challenge for next 24 hours of building!

Introduction to the challenge

Let's guide through the platform and Discord server! It's the best time to get all the answers for questions you may have

Hackathon Guide

Time is OVER

Creating a tool to make gmail more productive

Buntoshi

Sarthak

Gmail has a feature where you can fiilter emails automatically based on the sender and whether it has certain words or not. But these filters are not smart. What if the email can be understood and assigned a label before they are read by a human. That is what this tool does. It looks at the present emails and their labels and learns to classify newer incoming emails. That way you can decide which email requires your attention more and which doesn't even before reading it. 

How it works:
1) The user classifies their emails based on the content.
2) They activate the addon which can downloaded from Google Workspace Marketplace. 
3) After that, any important incoming mail will be labeled automatically.

The new email will be the input to Cohere's classify model which learns to classify using older human-labeled emails. 

Limitations:
The accuracy will be low in the beginning. It might wrongly label emails. But this can be easily corrected with human feedback. The user just has to relabel that particular email and their correction will be considered by the AI model next time. 

It has a number use cases : It can be used by Sales reps if their client is interested and reply quicker to those emails. Freelancers can prioritize on emails that ask for a quote. It can be used by customer succes teams to route queries to different departments and so on...

An add-on that classifies emails into labels based on its content

"Buntoshi team on Build your AI Startup Hackathon. Episode 3 Hackathon"

Team Idea

Submission