Our Model
A New Public Health Operating System
Khushi Baby’s operating model strengthens public health systems by closing the gap between data, insights, and action, particularly at the last mile of service delivery.
Data. Insights. Action.
Our model starts with one framework, run as a loop until it completes: data the system can trust, insights officials can act on, and action that reaches the household.
DATA
Our digital platforms make data collection simple and free of duplication for frontline health workers. Information captured on the ground is consolidated and made visible to the right stakeholders in the system.
INSIGHT
Our analytical infrastructure converts field data into clear, decision-ready insights that are presented in ways that make action straightforward for everyone in the system.
ACTION
Our health action centres enable data-driven decisions and improved health services to close the loop of healthcare to the last-mile communities. This can come from a change in policy, program, practice or partnership.
Two levers complete the loop.
We deploy a technology platform (CHIP) and an embedded process (Health Action Centers) to complete the Data–Insights–Action feedback loop.
The Technology Platform
CHIP
The Community Health Integrated Platform is the app in the frontline worker’s hand and the dashboard on the official’s desk. One family, one record, across every health program, feeding insights that become assigned tasks.
The embedded process
Health Action Centers
A lean, multidisciplinary unit inside the district health department. It triages the highest-risk cases, tasks callers and field teams, and turns every gap on the dashboard into an assigned, time-bound action.
How we implement it
Six steps, run in sequence, in every geography we serve.
01
Step one
Understanding On‑ground Needs
Every deployment begins by sitting with district and state officials to identify the most pressing health priority: maternal mortality, child malnutrition, TB, immunization gaps, or a disease no list has ever counted, like sickle cell in Nandurbar. We establish a baseline and define what closing the loop looks like here.
This takes time, field visits, and co-design with community health workers, who understand barriers to care that no dashboard will surface on its own. We map where the Data‑Insights‑Action cycle actually breaks in this district. The biggest pain point becomes the first requirement.
02
Step two
Building Infrastructure to Support the Process
Everything we build sits inside government, from day one to handover. Before a single data point is collected, we operationalize MoU terms and agree on entry and exit strategies with our government counterparts. We map partners, develop government budget lines for long-term ownership, and raise funds together with the department. A clear mandate to reduce duplicate data entry for health workers is a non-negotiable starting condition.
We take stock of the digital systems that already exist, integrate with them, and add only what is missing. And we establish the Health Action Center, the embedded unit that will coordinate the public health response once the data starts to speak.
What this looks like in practice
We sign five-year MOUs with state governments. The Rajasthan and Karnataka governments finance programmatic costs, including data plans, servers, and a bring-your-own-device policy for health workers. Karnataka selected Khushi Baby as its digital public health platform partner through a competitive bid and pilot process.
03
Step three
Effective Data Collection
Community health workers use CHIP to enumerate every household across the geography, establishing a live registry of named, located individuals that every health program can track over time. Longitudinal care modules then follow each person through pregnancy, immunization, nutrition, TB, and beyond. This is the substrate on which all insights and actions depend.
CHIP is one source among several. In every geography we hunt for the highest-signal data available: SNCU discharge registers in Nandurbar, ANMs’ own nominations of their riskiest pregnancies, and third-party datasets like travel time to the nearest facility and Poshan Tracker feeds. These often outperform routine systems that have been in place for years. We stitch them into one picture.
04
Step four
Decision‑Grade Insights
Field data becomes actionable through AI and GIS dashboards that show where high-risk populations are concentrated, where referral loops are breaking down, and which individuals need follow-up before their window closes. A vulnerability map narrows 952 villages to the 150 that need help first. Insights flow to officials at block, district, and state level, including health secretary and chief minister offices, and are directly linked to action scheduling.
Supportive supervision structures for community health workers, live data quality monitoring, and an inbound and outbound call center ensure the system does not stall at the insights layer.
05
Step five
Activating the Health Action Center
Now the Health Action Center established in step two goes to work. In districts identified by NITI Aayog and government as top-priority and aspirational geographies, a lean interdisciplinary unit deploys alongside the health department and runs a command center, drawing on the urgency of our COVID-19 war-room experience.
The HAC is the unit that coordinates the public health response, through policy, program, practice, or partner activation. It works through the government’s own health workers: triaging the highest-risk cases, tasking callers and field teams, and turning every gap on the dashboard into an assigned, time-bound action that is tracked, acted on, and measured.
06
Step six
Driving Change Beyond Action Units
Closing the loop at the district level creates evidence and tools that travel further, informing policy at state and national level and enabling partner organizations to reach communities Khushi Baby cannot.
Playbooks, datasets, and platform components are open-sourced, so partner NGOs and governments can deploy proven solutions without starting from scratch, and reach deeper into communities we cannot directly serve.
What we learn becomes public evidence: a randomized controlled trial showing a 12% improvement in full infant immunization, peer-reviewed publications, and $20M in government financing unlocked for state-led scale-up. CHIP data has shaped national TB surveillance strategy, Maharashtra’s tribal health policy, and district action plans for climate-health vulnerability.
Evidence is how one district’s work becomes every district’s baseline.
The model in action-NURTURE
Identifying and treating malnourishment in children
NURTURE is a white-labeled case-management module of CHIP, built at the request of Maharashtra’s Department of Women and Child Development. It solves a broken handoff: ICDS identifies severely malnourished children, the Health Department treats them, and for years no shared system connected the two.
42,000
SAM children tracked end to end
36
All Maharashtra districts
2,500+
ICDS supervisors trained
India's malnutrition crisis is often misdiagnosed as an identification problem. In reality, the systems to find malnourished children exist but the systems to treat them are fractured. Of approximately 100,000 SAM children identified annually in Maharashtra through the Poshan Tracker, only 10–15% are accounted for in treatment. The rest disappear from the system because no shared system exists to close the loop.
NURTURE is Khushi Baby's response. It is an end-to-end child nutrition and health tracking program that converts passive identification into an active feedback loop thereby closing the gap between a child being seen and a child receiving care.
Before building anything, Khushi Baby sat with supervisors, public health staff, and district officials to find where the system was breaking down. Children needed clinical verification and referral, but no one owned those steps, no record was shared between departments, and no one could see whether a child had received care.
NURTURE pulls SAM and MAM identification data directly from the national Poshan Tracker, eliminating duplicate data entry and giving ICDS supervisors a live, ready-to-action line list of children who need follow-up. Both departments now track the same child from identification at the Anganwadi Centre through treatment and recovery.
The dashboard shows where children drop out of the care continuum between identification, verification, referral, treatment, and recovery. Weekly state-level reviews use it to find bottlenecks and act. 42,000 SAM children are actively tracked across 95% of Maharashtra’s local administrative units.
ICDS supervisors use NURTURE to validate SAM status in real time, prioritise home visits for the most severe children, and confirm that referrals are completed and treatment begins. Over 2,500 supervisors have been trained. Their dashboard reviews have informed over half a dozen policy amendments by the Department of Women and Child Development, and the model is owned and operated by the Maharashtra government, designed for replication in Rajasthan and Karnataka.
