LEAP 2024 Hackathon 

Registration opens

Opening remarks by NewNative, discussing Lablab's mission and vision

NewNative - Brief Introduction to the Hackathon, Challenge, lablab.ai platform, and Discord server.

NewNative. Participants are encouraged to begin working on their projects.

Official Hackathon Kick-off - NewNative

It is a perfect time to connect with potential team members and energize with coffee.

1st Matchmaking Coffee session begins

Emmanuel Iriarte - Building your AI application in rapid prototyping

Workshop

2nd Matchmaking Coffee session begin

Participants resume working on their projects.

Harry Berg

How to launch your career in Ai & Data at supersonic speed.


How to Architect and deliver production-grade, high-speed computation and artificial intelligence software for real-time analytics

Workshop - Anton Alexander

Hacking Commences.

AI personalization in social media and marketing

Ash Barbour 

Workshop -  Jens Ingelsted - How to Pitch 

Hacking Commences

Lossless Data Compression based on new math theory & algorithm

Workshop - Claus Rosenberg Gotthard 

How AI is changing the way we build and operate software


Pawel Czech

End of submissions. Judging commences, and finalists are selected.

End of Closing Ceremony

Winner announcements at the main stage

Voice Enabled AI Assistant that will help digitally illiterate population to access web apps.

RAHA

Rubaiyat94

mumtasir2000

With the AI technologies growing in rocket speed, there is a large population globally getting no access to this AI diaspora. We want to include them with accessibility. Raha is designed to address the digital inclusion gap, using Large Multimodal Models (LMMs) to provide a voice interface for individuals without access to conventional AI technologies. It employs a model and tool-agnostic approach, ensuring compatibility with all desktop GUIs, including virtualized environments like Citrix and web interfaces.  Raha uses Open Adapt, which's auto-prompted methodology, derived from human demonstrations, ensures grounded agents, minimizing errors in task execution. With a focus on practicality, Raha represents a step forward in making AI more universally accessible.

There are 3 core stages, Voice Input Analysis, Foundation Models (e.g. GPT-4, ACT-1) which are powerful automation tools and OpenAdapt which connects Foundation Models to GUIs. 

Being tool agnostic, and an alternative approach to RPA, we are simplifying the whole digital journey for an individual with even digital illiteracy. Apart from this, enormous volumes of mental labor are wasted on repetitive GUI workflows, we see business scope in that too.

Raha is a voice enabled AI, powered by Large Action Models to enable Digital inclusion in AI diaspora. It's Model and Tool agnostic, can work with all types of desktop GUI. 

"RAHA team on LEAP 2024 Hackathon Hackathon"

Team Idea

Submission