Avani – Audio-Based Farm Insights for Better Policy
- Year2026
- CategoryImpact Achievers
- StageFinalist
Software Engineer
I applied for the GenAI Zürich Award because Avani stands for the kind of artificial intelligence we urgently need. Not AI that optimizes ads or automates convenience, but AI that restores trust in broken systems. In agriculture, decisions that affect millions of farmers are often made using flawed or fabricated data. Avani proves that a simple, voice-based AI system can listen at scale, earn farmer trust, and convert lived realities into credible insights for policy and climate action.
This project is driven by a clear belief: intelligence is only as valuable as the people it empowers. Farmers are not data points. They are decision-makers operating under uncertainty, climate volatility, and market instability. When their realities are accurately captured, policies improve, waste reduces, and resilience increases. We have demonstrated that this model works across states and crops. Now the challenge is to strengthen its technical depth, ethical foundations, and long-term scalability.
The GenAI Zürich Award represents more than recognition. It is an opportunity to subject Avani to rigorous scrutiny, to learn from leaders shaping responsible generative AI, and to ensure that what we are building can mature into public infrastructure. We have built proof. What drives us now is permanence, scale, and dignity in how technology serves those who feed the world.
Sahil Kale
Software Engineer
Agricultural policymaking is disconnected from ground-level reality. The process of ground-level data collection about plantation and harvest currently followed by agencies is ridden with challenges, leading to poor quality of data collected. This flawed and skewed data is used to decide policies and design schemes for farmers, causing a poor policy framework. This creates a structural disconnect between farmers’ lived realities and policymakers’ assumptions, leading to distorted schemes, poor market interventions, and systemic inefficiencies.
Avani addresses this root cause through an AI-powered, audio-based data collection and analysis system. Farmers receive automated calls in their regional language at predefined intervals and respond to structured questions about crop conditions, inputs, water access, labor, and production constraints. The responses are recorded and processed using natural language processing pipelines that transcribe, translate, segment, and extract structured insights. Both qualitative and quantitative signals are analyzed to generate actionable intelligence for government agencies and agrarian stakeholders.
The system is currently deployed across four Indian states, serving over 1,000 farmers across 30 plus crops with plans to expand in China and Peru and serving over 50 crop types by 2027. Avani has seen a high satisfaction rate among both parties: 92% of farmers agreed that their voices were being heard; 82% of decision-makers felt insights provided by the system were useful. By improving data fidelity at the source, Avani enables evidence-based policymaking, reduces information asymmetry, and strengthens environmental, economic, and social sustainability in agriculture.
Shape the future of GenAI in Switzerland and beyond
Join us for the next edition of GenAI Zürich to learn about latest developments in the field, network with like-minded professionals and gain deeper insights into GenAI applications reshaping both our professional and personal lives.