Loading Tech4Bharat…
01 / FIND YOUR CHALLENGE
Tech4Bharat is the hackathon of the Global Accelerator Vision Summit 2026, run as the JanAI Mass Entrepreneurship Challenge.
Student teams take on twenty real problems faced by small enterprises and the people they serve across Bharat, organised under eight sectoral themes.
Each theme groups two to four problem statements. The tag on every statement shows whether it calls for an Agentic AI system or an Applied AI system.
Unlocking the value of the largest asset most rural families own.
Information on housing schemes, subsidies, finance, approvals, construction costs and contractors is scattered across offices and people. Applications stall on missing or inconsistent documents, and families depend on middlemen to move a process that can run for many months.
Ownership, inheritance, mutation and boundary records are hard to understand for families whose land is their main asset. Rural property rarely has a credible valuation, so it cannot be used as collateral and lenders stay away.
JanAI Mass Entrepreneurship Challenge · 8 themes | 20 problem statements
Tech4Bharat is the hackathon of the Global Accelerator Vision Summit 2026, run as the JanAI Mass Entrepreneurship Challenge. It is an initiative of Karnataka Arya Vysya Mahasabha, in partnership with JanAI and in collaboration with RV College of Engineering. Student teams take on twenty real problems faced by small enterprises and the people they serve across Bharat, organised under eight sectoral themes.
The statements draw on the field inventory of rural MSME problems compiled by Madan Padaki and on the JanAI rural agentic AI problem set.
Download the brochure (PDF, 22 pages)Each statement describes a problem and leaves the solution entirely to you. We expect teams to take a statement into the field, sit with the enterprises and families who live with it, test what is written here against what they find, and only then define the solution they want to build on the hackathon platform.
Every statement follows the same layout
Most statements are marked Agentic AI. These call for a system that pursues a goal over many steps on behalf of its user: it plans, acts in external systems, keeps memory of the case, recovers when something fails and seeks human approval before any action involving money, legal commitment or personal data.
Sandboxed or simulated integrations are allowed as long as they are disclosed, and the bar is a system that acts for its user, well beyond answering questions.
Statements marked Applied AI Systems put the depth of the model, data or optimisation at the core, whether that is vision on commodity hardware, speech in Indian languages, forecasting under uncertainty, risk modelling, optimisation or market design. These are judged on measured performance.
Submissions are judged on two dimensions. The technical dimension covers architecture, quality of AI, a working prototype and innovation, while the entrepreneurial dimension covers problem relevance, market, business model, scalability and job creation. Final criteria and weightages are published with the registration details.
The people in these statements speak their own languages and often mix them, many have limited literacy, most use entry-level Android phones, and connectivity comes and goes. Solutions are expected to be voice-first and vernacular, to hold up in those conditions, and to offer an assisted path through a JanAI Café wherever documents, authentication or physical presence are needed. Consent is explicit, data collection stays limited to what the task needs, and mock integrations are clearly labelled wherever public or banking systems lack open interfaces.