KissanAI is a primary example of vertical agent orchestration. They are moving the agricultural sector from passive information retrieval (chatbots) to active operational agents through their Dhenu Agentic Platform. Their agents are designed to handle specific, high-stakes agricultural workflows including pest management, input procurement, and supply chain coordination.
In the broader agent stack, KissanAI operates at both the model layer and the application layer. By releasing the Dhenu LLM series, they provide the domain-specific "brains" that other developers can use to build agricultural agents. Their emphasis on voice-first interfaces and high-accuracy multilingual support addresses a critical frontier for agent deployment: making autonomous systems accessible to non-technical users in offline or rural environments. They are effectively championing the "Agentic OS" for the agricultural industry.
KissanAI is an agricultural technology company that builds domain-specific artificial intelligence for the global farming ecosystem. Based in Surat, India, and founded in 2019 by Pratik Desai and Chintan P., the company began as Titodi Infotech before focusing on the potential of large language models to solve the persistent digitization gap in agriculture. The company is known for launching KissanGPT, a viral AI chatbot for farmers, which has since evolved into a broad platform of agents and foundational models designed to orchestrate complex farming operations.
While many AI startups apply horizontal models like GPT-4 to vertical problems, KissanAI takes an inverse approach. They maintain that general-purpose models lack the nuance required for high-stakes agricultural decisions, such as pest management or input optimization. To address this, they developed Dhenu, a series of open-source agricultural LLMs. The Dhenu 2.0 model, built on Meta's Llama 3, is trained on approximately 1.5 million agriculture-specific instructions across 4,000 categories. These models are published openly on HuggingFace, making KissanAI a central contributor to the open-source agricultural AI research community.
One of the most significant technical hurdles in agricultural AI is the interface. In many of KissanAI's primary markets, farmers rely on voice communication and local dialects rather than text. KissanAI has focused heavily on voice accuracy, achieving 95% accuracy for agricultural conversations in Hindi. Their Kissan Agent is a voice-enabled assistant that provides personalized advisory services, translating agronomic science into actionable steps for smallholder farmers. This focus on accessibility has allowed them to reach over 100,000 farmers across Asia and Latin America within a few years of launch.
For enterprise clients, the company offers the Dhenu Agentic Platform, formerly known as AgriCopilot. This system is designed for agribusinesses and government institutions to automate advisory delivery and supply chain intelligence. By moving from simple Q&A to an "agentic" model, the platform can actively coordinate tasks like procurement and market access. The company is currently scaling this infrastructure to serve more than 4 million farmers through partnerships with major agribusinesses.
KissanAI occupies a unique middle ground between a research lab and an enterprise software provider. They are a founding member of the AI AgriBench alliance and participate in major industry programs including NVIDIA Inception and the Microsoft AI First Mover initiative. Their work with the United Nations Development Programme (UNDP) further illustrates their role as a bridge between high-end AI research and practical, field-level applications.
As they expand into North American markets, KissanAI is adapting its stack for large-scale commercial farming. This expansion involves "Input Intelligence" tools that combine agronomic science with decision support for precision agriculture. Whether serving a smallholder in Gujarat or a large-scale producer in the Midwest, the underlying thesis remains the same: agriculture requires a specialized AI layer that understands the physical reality of the farm as deeply as it understands language.
Voice-enabled AI assistant providing personalized crop advisory and market insights.
Open-source, agriculture-specific large language models trained on domain data.
KissanAI is hiring.