
The evolution of Voice AI - from conversational tools to intelligent, action-oriented interfaces - has opened the gates for businesses to rethink customer engagement, streamline operations and unlock new opportunities for growth. From helping field teams access enterprise information hands-free to enabling transactions through spoken instructions, Voice AI helps businesses operate, interact with customers and deliver outcomes.
To encourage the development of practical solutions that leverage Voice AI, technology partner Google Cloud, YourStory, and Hack2skill have launched Build with Gemini โ Shape What's Next, a fully virtual hackathon aimed at driving hands-on adoption of the Gemini stack among working professionals across India.
The hackathon invites developers, AI/ML engineers, data scientists, and technical founders to identify business challenges and develop functional voice AI prototypes using Gemini.
Participants can register individually or in teams of up to four members. Students and non-professionals are not eligible for this edition.
Registrations and idea submissions are open till December 3, 2026. Participants must submit an idea addressing one of four problem statements. Shortlisted teams will receive $100 in Google Cloud credits to develop their prototypes and submit a one-to-two-minute demo video.
Four problem statements
The hackathon focuses on building Voice AI solutions that go beyond answering questions to retrieving information, completing tasks, and taking action through connected systems.
Participants can choose from four problem statements.
1. Voice-powered enterprise knowledge and decision agent
Participants must build a real-time voice agent that helps employees retrieve information from multiple enterprise sources, including documents, support tickets, project wikis, and chats.
The agent should support regional languages and code-switched speech, retrieve information across languages and provide answers grounded in enterprise data. It should also carry out actions through connected systems, such as raising tickets, checking inventory, or updating records.
2. Voice-driven agentic workflow automation
This challenge focuses on using spoken instructions to complete business tasks without requiring users to navigate forms or websites.
Participants must build voice-first agentic workflows that convert multilingual spoken instructions into actions across APIs, ledgers, databases, or inventory systems.
The solution should coordinate multiple steps, handle exceptions, and involve a human when approval or judgment is required. Potential applications include logging field visits, recording deliveries, booking services, and filing claims.
3. Customer & frontline voice agents
Participants must develop a real-time voice agent that interacts directly with customers or assists frontline employees during live conversations.
The agent should understand intent, context, and sentiment; retrieve relevant information and complete tasks such as booking a slot, checking orders or raising support tickets.
It should work across languages and include a human hand-off mechanism for situations requiring judgment.
4. Open innovation: Real-world voice AI with Gemini
This category allows participants to identify an organizational challenge or opportunity in any industry and develop a voice AI solution.
The agent must use Gemini to understand and reason over spoken input, including Indic languages and code-switched speech, and connect that input to actions through APIs, ledgers, databases or inventory systems.
Across all four categories, participants are expected to demonstrate natural, low-latency voice interactions; regional-language capabilities; the ability to handle interruptions and background noise; and the execution of backend actions.
Prototypes should also demonstrate measurable value through factors such as time saved, tasks completed, improved accuracy, language reach or cost per interaction.
From idea submission to prototype
The hackathon follows a seven-step process, beginning with free registration and idea submission.
Shortlisted teams will receive Google Cloud credits and move into the development phase, where they will build prototypes using Gemini and submit a demo video.
Participants will have access to industry mentorship, office hours and guided sessions led by practitioners from Google Cloud and Hack2skill.
Submissions will be evaluated in two stages. The process includes AI-assisted and manual screening of ideas, technical pre-screening, rubric-based scoring and a final jury round involving Google Cloud subject matter experts.
The evaluation criteria include the use case and its business impact, effective use of Gemini Enterprise capabilities, product quality and execution, innovation, presentation and user-generated content.
Rs 8 lakh prize pool
The hackathon offers a total prize pool of Rs 8 lakh, with eight winning teams selected from a single national pool.
- Winner: Rs 3 lakh
- First runner-up: Rs 2 lakh
- Second runner-up: Rs 1.25 lakh
- Five consolation awards: Rs 35,000 each
All shortlisted teams will receive $100 in Google Cloud credits to support prototype development.
Every team that submits a valid prototype will receive an e-certificate. Selected participants may also have their work featured across YourStory, and Hack2skill channels.
Working professionals interested in building Voice AI solutions can register and submit their ideas until December 3, 2026.


Comment
0Upholding journalistic discourse. No registration required.
No comments yet. Share your perspective!