Last month, over 270 skilled participants gathered in Bengaluru for the first-ever Llama Hackathon in India, hosted by Reskilll in collaboration with Meta. The event attracted more than 1,500 applications, with an impressive 350 teams submitting ideas in the pre-selection round, from which 95 teams were invited to compete. With support from more than 25 mentors and six distinguished jury members, these talented teams embarked on a 30-hour journey to create AI solutions tackling real-world challenges, leveraging Meta’s Llama 3 models and WhatsApp APIs.

The event saw support from partners like AWS, Qdrant, Tune AI, Sarvam, and LlamaIndex. Over 25 mentors from diverse industries contributed their expertise, guiding participants to bring their ideas to life and push the boundaries of AI innovation.

Hackathon Highlights:

• 1,500+ applications
• 95 teams invited
• 270+ participants
• 75+ projects completed
• Three themes: AI for Societal Good, Build on WhatsApp, and AI on Edge
• Total prize money: $7,000 USD
• Fueled by creativity and caffeine, our participants powered through the event with over 800+ cans of energy drinks and 80+ boxes of coffee and tea.

First Place: CurePharma AI (Theme: AI for Societal Good)
Prize awarded: $3,000 USD

CurePharma AI is an innovative healthcare solution designed to offer personalized medication information in regional languages, addressing accessibility issues in linguistically diverse, rural, and literacy-challenged areas. By integrating WhatsApp, it offers an easy-to-use interface where users can scan prescriptions, access local language translations, and even order medications via e-pharmacies. Through Llama’s advanced language capabilities, CurePharma AI bridges the gap in healthcare information accessibility, creating a seamless experience for users across demographics.

Second Place: CivicFix (Theme: Build on WhatsApp)
Prize Awarded: $2,000 USD

CivicFix provides an AI-powered bot on WhatsApp to help citizens report local issues like potholes, waterlogging, and streetlight outages through voice notes, photos, or simple descriptions. The bot classifies, prioritizes, and directs reports to relevant local authorities for prompt action. CivicFix streamlines the process, making it easy for users and efficient for authorities, who receive AI-prioritized insights based on frequency, severity, and location. Over time, CivicFix will generate comprehensive summaries for IAS officers, driving more efficient and proactive governance.

Third Place: Evalssment (Theme: AI for Societal Good)
Prize awarded: $1,000 USD

Evalssment aims to reduce educators’ workload in grading and assignment creation through an AI-driven solution. The tool leverages Llama models to grade assignments and provide individualized feedback, allowing teachers to set personalized grading criteria and generate assignments with one click. Evalssment brings transparency, efficiency, and bias-free grading to the education sector, freeing up educators’ time for more hands-on mentorship, ultimately enriching the student experience.

Best All-Women Team Award: Aarogya Assist from SheBuilds (Theme: Build on WhatsApp)
Prize awarded: $1,000 USD

Aarogya Assist is a healthcare aggregator designed to simplify access to essential healthcare services through WhatsApp. From medication inquiries and lab report access to insurance information, Aarogya Assist provides a one-stop platform accessible in the user’s preferred language. This inclusive solution leverages Llama’s language modeling and WhatsApp’s widespread accessibility to address healthcare needs conveniently and equitably.

Other noteworthy ideas:

SPAM - LX Reasoning Engine (AI for Societal Good)
A text-to-ASL translation tool using LLMs, STM, and reasoning engines to create culturally appropriate translations. This tool aims to expand accessibility and communication for the Deaf community. The solution deploys the LX Reasoning Engine to define and leverages many of the critical components necessary for effective text-to-ASL translation leading to more accurate and culturally appropriate translations by integrating advanced technologies like LLMs, STM, LTM, and reasoning engines.

KaryaAI (AI for Societal Good)
Using a multi-agent approach, KaryaAI creates story-based curricula that are culturally diverse and contextually relevant. Llama enables KaryaAI to design customized learning experiences for data annotators, catering to diverse regions across India.

Auto Bot (AI on Edge)
Circuit Bot is an edge-based personal assistant for home automation that integrates ASR (Automatic Speech Recognition) and a Llama model to interpret user voice commands. This system allows users to control BLE (Bluetooth Low Energy) appliances directly, eliminating the need for internet connectivity. The primary goal is to provide localized, low-latency control over home appliances with flexibility for users to code, customize, and manage appliances through both the assistant and an accompanying app.

Benjamin Button (AI on Edge)
This team introduced a novel model compression technique tailored to specific tasks, enabling developers to select a model size suited for their data. Benjamin Button provides flexibility in AI solutions, especially useful for edge devices.

KPMG - Data Citizen -Grant Guru (AI for Societal Good)
Grant Guru is an AI solution that helps non-profits discover and apply for grants, helping them find funding opportunities and reducing the time spent on applications, so they can focus on making a difference.

HeyHomie (Build on WhatsApp)
HeyHomie offers a SaaS platform for homepreneurs to grow their businesses on WhatsApp. Their AI-powered Ads Studio enables small business owners to create and manage click-to-WhatsApp ads, helping them connect with customers and expand their reach.

Our commitment to an open AI ecosystem was evident as developers used the latest Llama models to create impactful projects. Attendees marveled at the accessibility and versatility that Llama offers, making it possible to create complex, impactful solutions in a short time. From empowering civic engagement to enhancing healthcare accessibility, this hackathon was a showcase of innovation.

Thank you to everyone who participated in and contributed to the success of this inaugural AI Hackathon with Llama in India. We can’t wait to see what this community will continue to build with Llama’s tools and capabilities.