Yale Summer Trainee Positions MAHILA-RW

Host and research team

Host: Khushi Baby Association, Udaipur, in collaboration with the Yale School of Public Health (Yale Institute for Global Health, Hecht Global Health Faculty Award).

Principal investigators: Prof. Bhramar Mukherjee (Yale School of Public Health) and Dr. Ruchit Nagar (Khushi Baby).

Each trainee leads a defined component during the start of full enrollment and works with the Khushi Baby data science and field teams, the YSPH biostatistics group led by Prof. Mukherjee, and two Indian student interns working on model adaptation and fairness evaluation.

The project

MAHILA (Maternal Anemia Hemoglobin Imaging for Last-mile Assessment) is a smartphone AI tool that scores conjunctival (inner-eyelid) images to help health workers decide which pregnant women need a hemoglobin test first.

The current multi-site model (manuscript in preparation) was developed on 5,385 pregnant women and more than 32,000 conjunctival images from five sites in four Indian states, against venous CBC hemoglobin, with leave-one-site-out, temporal and pre-registered external validation. Its signal is strongest for severe anemia (Hb <7 g/dL) and weaker for moderate anemia, and referral cut-offs must be set locally.

In 2027, the study moves to real health camps: ASHA community health workers in Nandurbar (Maharashtra) and Udaipur (Rajasthan) will screen a target of 3,000 pregnant women, each generating a paired ASHA-image + nurse-image + CBC reference triad.

Two questions remain open: how the model performs when operated by ASHAs in variable community conditions, and whether the consent-to-result workflow is feasible in real camps. Two validation datasets, controlled facility and real-world community, both with laboratory reference standards, support adapting a locked base model into a real-world model (MAHILA-RW).

Findings feed a policy brief and implementation playbook for India’s National Health Mission, ICMR, and state health departments on integrating AI screening into the Anemia Mukt Bharat program.

The positions

Requirements for both positions

• Comfort working in rural and tribal field settings, in heat and on long camp days.
• Ability to work closely with community health workers with respect for their time and expertise.
• Disciplined documentation and version control.
• Eligibility to travel to India for the summer and obtain the appropriate Indian visa.
• Completion of Yale human subjects (CITI) training before departure.

What you gain

• Co-authorship on manuscripts and the policy brief, per ICMJE criteria.
• Mentorship from Prof. Bhramar Mukherjee (YSPH Biostatistics), Dr. Ruchit Nagar, and the Khushi Baby data science and field teams.
• Hands-on work with a pregnancy-focused smartphone anemia imaging dataset, with two validation sets (facility and community) that both have laboratory reference standards.
• Direct exposure to policy audiences: NHM, ICMR, and the state health departments of Rajasthan and Maharashtra.
• Field experience with a nonprofit that runs digital health systems for state governments, across community health workers serving tribal and rural districts.
• For MPH students: the placement may meet the YSPH Applied Practice Experience requirement; confirm with your department.

Terms

• Duration: approximately 3 months, summer 2027 (June–August; dates set with each trainee).
• Location: Khushi Baby office, Udaipur, Rajasthan, with field visits to study sites in Udaipur district and Nandurbar, Maharashtra.
• Funding: stipend, round-trip international airfare, and in-country lodging and subsistence for the placement.
• Supervision: weekly virtual check-ins with Prof. Mukherjee’s group and day-to-day supervision by Khushi Baby in Udaipur.
• Trainees follow Yale international travel registration and safety requirements and the study’s approved IRB protocols.