A recent article by Digital Green CEO Nidhi Bhasin explores how India’s next phase of digital agriculture will depend less on scaling generic advisory platforms and more on leveraging AI to deliver contextual and actionable support for smallholder farmers.
In a country where agriculture employs more than 40% of the workforce, agritech startups are increasingly delivering weather, pest and market intelligence at scale. The piece argues that many first-generation digital advisory systems struggled because recommendations were often too generic or disconnected from local realities, limiting on-farm impact and behavioural adoption.
AI-driven models are now being used to personalise advisory around local weather conditions, crop cycles and farmer-specific needs. Digital Green’s Farmer.Chat platform, reaches around 7 million farmers across India and Africa, with approximately 60% of users reportedly taking action based on recommendations received through the platform. The article also highlights the World Economic Forum-backed Saagu Baagu initiative in Telangana, which uses AI and data-driven advisory to support chilli farmers through more localised recommendations.
Image credit: Digital Green
The piece argues that AI’s greatest contribution may not be making advisory more localised and relevant, but fundamentally changing the economics of agricultural extension. According to Digital Green, traditional extension services cost around $35 per farmer, compared to roughly $3.5 for video-based advisory, while GenAI-powered systems could reduce costs to as little as $0.35 per farmer.
At the same time, inclusion remains a challenge. Women account for approximately 45% of Farmer.Chat users, yet barriers around smartphone access, digital literacy, and participation in agricultural decision-making continue to limit adoption. The article warns that without stronger coordination between agritech platforms, public digital infrastructure, and government-led inclusion efforts, AI could reinforce rather than reduce existing inequalities.