Amul’s new AI platform begins with an advantage most agricultural technology companies spend years trying to create: five decades of cooperative records tied to real farmers, animals, veterinary care, and milk production.
The Indian dairy cooperative has launched Sarlaben, an assistant intended to give 3.6 million women milk producers personalised guidance at any time. It is available through the Amul Farmer mobile application, already downloaded by more than one million users, and by voice call for people using feature phones or landlines.
The ambition is not simply to answer general farming questions. Integration with Amul’s milk-collection and livestock systems allows cattle-specific advice grounded in individual history.
Cooperative Infrastructure Makes Personalisation Possible
Amul’s digital backbone manages more than two billion milk-procurement transactions annually. Its data includes veterinary records from over 1,200 doctors covering nearly 30 million cattle, roughly seven million artificial-insemination procedures each year, satellite imagery for fodder mapping, and a cattle census conducted every five years.
Each animal has a unique identifier connected with feed, disease, and milking information. That enables an assistant to move from generic advice towards recommendations about a particular animal’s health, breeding, nutrition, or production.
The cooperative structure is central. Amul did not collect data solely to train a model; the records accumulated while operating milk procurement, veterinary support, and breeding programmes across thousands of villages. AI is a new access layer over an existing institution.
This foundation can support predictive disease detection, oestrus tracking, feed formulation, and local weather-risk advice. Its quality still depends on timely and accurate records, including observations that may be difficult to capture consistently in rural environments.
Voice and Language Determine Who Benefits
Sarlaben launches in Gujarati and uses India’s Bhashini multilingual framework, creating the possibility of expansion to 20 Indian languages. Language support is essential in a network that extends across villages and local dialects.
Voice access may be even more important than translation. The farmers with the greatest information gap may not be comfortable with smartphones or text interfaces. A midnight call about a sick animal needs to work on available equipment, understand local speech, and identify when automated guidance is insufficient.
Amul says farmers outside its cooperative can access general dairy and husbandry information, but personalisation relies on the records available inside its systems.
The platform should track reach across device types, languages, age groups, and locations—not only total queries. If smartphone users receive most of the benefit, the deployment could widen the information divide it intends to close.
Veterinary escalation must also remain clear. Advice about urgent symptoms should route farmers towards professional help rather than presenting a confident answer from incomplete information.
Productivity Is the Decisive Test
India produced 347.87 million tonnes of milk during 2024–25, more than twice the US total cited by the source, yet its yield per animal remains comparatively low. Small herds, feed quality, limited rural veterinary access, and uneven knowledge contribute to that gap.
Amul’s Gujarat network spans more than 18,600 villages and supplies around 35 million litres daily. Even modest improvements in animal health or yield could influence income at enormous scale.
The strongest evaluation will therefore measure outcomes beyond application use: earlier disease treatment, feed efficiency, conception, milk yield, veterinary referrals, and farmer income. Advice must be compared with what would have happened without it.
The assistant also needs feedback when a recommendation fails or local conditions differ from its data. Farmers and veterinarians should be participants in improving the system, not passive recipients.
Amul AI is a compelling example of technology built upon institutional memory. Its scale comes from the cooperative relationships that already connect transactions, animals, and services.
Whether it becomes a new “White Revolution” will depend on the last mile. The model may contain decades of evidence, but its value is realised only when a farmer can hear relevant guidance, trust it, act in time, and see a healthier animal or stronger livelihood as a result.
