The Wadhwani Institute for Artificial Intelligence was formed in 2018 to build AI solutions that benefit underserved populations in developing countries. We partner with government bodies and global nonprofits to ensure that our innovations will improve the lives of those who need them the most.

Currently we are building AI-based solutions in the agriculture and health domains, such as pest management for cotton farms, maternal, newborn and child health, tuberculosis and COVID19.

Pest Management  
Our solution is being developed to help reduce crop losses through integrated pest management in cotton farms.

We are creating technologies towards: 

  • Automated reading of TB LPA test results.
  • Providing differentiated interventions for TB patients by applying critical resources appropriately. We predict a score for patients most likely to prospectively not adhere to the treatment regimen, when they come in for their first treatment.

Newborn Anthropometry  
We are creating a smartphone-based anthropometry technology, which will allow frontline workers to track baby weight in rural homes and hospital settings.

We provided epidemiological forecasting for COVID-19 to assist in resource planning. We also conducted research to identify signature patterns for COVID-19 using cough sounds.


Systems that are sustainable and scalable can transform the lives of billions of people. We believe that using AI to solve problems at the bottom of the pyramid can lead to a better, more equitable world for all.

“We are in an age where efforts to achieve the UN Sustainable Development Goals are accompanied by a revolutionary explosion of digitally available data and the penetration of internet-enabled smartphones into previously inaccessible rural locales. AI technology is the natural tool for leveraging this vast scaleup in the quantity and breadth of data into actionable machine learning models that direct on-the-ground interventions for underserved populations. “It is expected that while we are in the midst of this accelerated growth, data sources will be unstable, incomplete, and erroneous, presenting a key challenge in developing AI models. At the other end of the tech pipeline, AI solutions should be designed to facilitate delivery to the last mile user without significantly perturbing existing public systems.”

– Alpan Raval, Chief Scientist, AI/ML


Effective solutions start with a systematic, long-term commitment to use AI for good. This involves working closely with partner organizations to create a conducive environment for defining problem areas, and developing and implementing solutions – a process we define as ‘AI-readiness’.


This refers to a partner organization’s ability to create and use AI solutions to achieve certain benefits. 

Ability: The capacity to engage in AI problem definition and to support AI solutions sustainably.

Create: Once a specific problem is defined, the actions (including technologies and methods) that may be used to build AI solutions.

Use: Activities involved in effective application of the solutions

Benefits: Anticipated benefits at the beginning of the project should match benefits accrued once it is implemented. It includes course correction and actions to modify a partner’s Ability.


We have identified the key filters to measure the AI-readiness of an organisation or a specific project.

For our partners, AI-readiness provides a link between long-term activities to be performed at the organization-level and precise benefits accruing from the implementation of AI in specific projects.