Sanchalak: Revolutionizing Rural Welfare Access
Voice-First AI for Inclusive Governance
Team Members:
Diya Ravishankar, Mohit Kumar, Raazi Faisal, Yuganshu Kumar
Mentors:
Personal Mentor: Pradeep Kumar
Annam Mentors: Soma Dhavala, Jiger Doshi, Sharesth Vij
Why We Chose This Problem Statement
At the heart of our project lies a stark and persistent issue — millions of farmers and rural citizens across India remain excluded from accessing welfare schemes that are legally and morally theirs. While the government has built a robust ecosystem of subsidies, support programs, and social benefits, the last-mile delivery is riddled with friction.
The Core Problem: Procedural complexity, language barriers, digital illiteracy, and lack of personalized assistance often mean that these schemes exist on paper but not in practice for many rural households.
We selected this problem statement because it speaks directly to the intersection of technology, governance, and social equity. With the rise of smartphone penetration in rural India and a growing openness to vernacular and voice-first technologies, the timing felt right. We saw an opportunity to bridge this critical access gap using conversational AI — not just as a fancy tool, but as an equalizer.
Our Mission: We weren’t just solving a tech problem; we were addressing a human rights issue. Our goal with AnnamAI was to rethink the way welfare services are delivered — shifting the paradigm from complicated forms and government portals to intuitive voice-based conversations that meet users where they are.
The problem is deeply relevant, incredibly high-impact, and systemically under-addressed, making it an ideal focus for an innovation-first hackathon.
Our Inspiration to Participate in the Hackathon
This wasn’t just another coding competition for us — it was a mission-driven opportunity. The hackathon’s theme of tackling “real Bharat problems” struck a chord with our entire team. Too often, hackathons orbit around urban conveniences — building yet another food delivery or ride-sharing solution. This event was refreshingly different. It challenged us to take our skills and apply them to areas that are typically underserved by mainstream innovation.
We were inspired by the potential of AI — especially multilingual and conversational AI — to make a real, lasting difference in the lives of people who have historically been left out of the digital transformation narrative. The idea that we could prototype something that would directly uplift rural livelihoods was incredibly motivating.
Our Vision: We wanted to prove that AI doesn’t have to be limited to elite urban use cases — it can be inclusive, empathetic, and grounded in everyday realities.
What It Was Like to Participate in a Hackathon Tackling Grassroots Realities
Participating in this hackathon was an eye-opening experience, one that redefined how we think about user-centric design. Unlike typical hackathons that emphasize speed over substance, this one forced us to slow down and engage deeply with the problem space.
We had to step out of the code and into the shoes of our users — smallholder farmers, elderly citizens, women in self-help groups — people for whom technology isn’t second nature. We conducted interviews with field workers and agri-extension officers, pored over complex government scheme documents, and worked to demystify bureaucratic language into something conversational, local, and actionable.
Design Challenge: Designing for low-connectivity areas, low-literacy populations, and vernacular speech inputs demanded more than just technical acumen — it required empathy, cultural nuance, and humility.
It was an incredibly rewarding challenge. For once, we weren’t building a prototype just to impress judges — we were building something that could be deployed on the ground, make a difference, and scale with purpose.
Challenges We Faced & The Road Ahead
The journey wasn’t without hurdles. Some of our biggest technical and design challenges included:
Major Technical Challenges:
Multilingual Voice Input Handling
Processing noisy, accent-rich speech in regional languages and accurately transcribing it posed a major challenge. We had to experiment with multiple ASR (Automatic Speech Recognition) backends and tune them for dialectal variations.
Structuring Unstructured Data
Government schemes are often listed in unstandardized formats across different websites and PDFs. Making this data machine-readable and contextually retrievable for our chatbot required custom scraping, cleaning, and structuring pipelines.
Maintaining Conversational Context
Designing an AI system that could remember previous user inputs, clarify ambiguities, and adapt its recommendations in a flow — all while keeping the interaction voice-first and latency-free — was a complex task.
These challenges, however, were instrumental in shaping the roadmap for AnnamAI. We now see immense potential to evolve the platform beyond just scheme discovery and into a comprehensive citizen support agent.
Future Vision
In the future, we envision:
🔗 Aadhaar Integration
Integration with Aadhaar-linked data to auto-fetch eligibility credentials and minimize manual input.
📱 Voice-Based Document Submission
Voice-based document submission via WhatsApp or SMS for truly paperless applications.
🌐 Comprehensive Citizen Services
Expansion into adjacent citizen services like old-age pensions, crop insurance, health subsidies, and digital literacy assistance.
Conclusion: A Prototype with Purpose
At its core, AnnamAI is more than a tech demo — it’s a step toward democratizing access, simplifying governance, and building inclusive digital infrastructure.
Key Insight: Participating in this hackathon reminded us that the real power of technology lies not in complexity, but in accessibility. We walked away not just with a working prototype, but with a renewed belief in the power of AI to drive meaningful, grassroots change.
We’re just getting started.
About Sanchalak
Sanchalak represents our commitment to building technology that serves everyone, especially those who have been traditionally underserved by digital innovation. Through voice-first AI, we’re making government welfare schemes accessible to every citizen, regardless of literacy level or technical expertise.