#servicedesign

Field notes are messy. Voices overlap. Contradictions appear. Patterns hide in plain sight.

After days of interviews, system mapping, and stakeholder conversations around e-Sanjeevani in Meghalaya, we reached the stage where everything needed to slow down.

The next step was about capture sheets.

For every interaction, we created structured capture sheets, not just to record what was said, but to decode what was meant. Each sheet helped us extract:

Who was involved

What their role in the system really is (not just on paper)

Pain points vs workarounds

Emotional signals (frustration, pride, hesitation, indifference)

Gaps between policy and practice

Some categories we focused on, for organising our data were:

Verbatim quotes

One month from now, we’d like to remember: 

Unusual / interesting observations we made

Initial thoughts/ inspirations for concepts

Patterns started emerging: • Repeated bottlenecks across different districts • Invisible labor carried by frontline workers • Digital constraints masquerading as “user error” • System-level delays that get absorbed at the human level

The most interesting part? The contradictions.

What the system assumes versus what the dashboard shows versus what the ground reality actually is.

Analysis, in this phase, wasn’t about numbers alone. It was about triangulation — aligning interview narratives, journey maps, consultation data patterns, and stakeholder structures to see where friction accumulates.

Putting the pieces together is less like solving a puzzle and more like tuning an instrument. You adjust one layer, and the whole system resonates differently. Design decisions are not made from empathy alone they are made from structured empathy. And tools like these are where structure begins.

If you’ve worked within layered public systems, especially in rural or resource-constrained contexts, we would love to hear your perspective. You can write to us at [email protected]. Until next time!

After mapping the structure of Meghalaya’s public health system, we zoomed in further to the people inside it. When we began listing stakeholders, the obvious ones came first: Patients, Doctors, Nurses 

But the map kept expanding.

IT support teams who troubleshoot failing connections. Pharmacists who interpret digital prescriptions. Supply chain actors who determine whether prescribed medicines are actually available. NHM officials managing policy and monitoring. Data operators entering consultation records. Private practitioners who remain outside the platform. Even informal providers - the local healers who compete with it.

Each actor influences the system by influencing the interactions that they bring in even if they are not visible on the screen.

The User Journey: It’s Not Just a Call

We mapped the user journeys of the major stakeholders - the doctors, the MLHPs and the patients. What looks like a 2-minute consultation is actually a multi-actor relay with not just the patient as the user but also the nurse, doctor and system operators. Designing telemedicine, then, is not about optimizing a video call. It is about aligning incentives, capabilities, and constraints across a distributed network of actors.

The more we map, the clearer it becomes - healthcare journeys are non-linear, and a lot more goes into providing healthcare that what is seen. Specifically, in the case of a state like Meghalaya, where terrain, seasons and travel are big constraints, healthcare is beyond just meeting with a professional. These journeys are navigated through availability, trust and infrastructure.

If you’ve worked within layered public systems, especially in rural or resource-constrained contexts, we would love to hear your perspective. You can write to us at [email protected]. Until next time!

As part of our ongoing study of e-Sanjeevani, we stepped back from individual consultations and began mapping the larger public health system of Meghalaya.

What emerged was not just a platform, but a layered, distributed structure.

At the base are Sub-Centres and Health & Wellness Centres, often staffed by MLHPs and nurses. These are the first points of contact in rural areas. Above them sit Primary Health Centres (PHCs), then Community Health Centres (CHCs), and District Hospitals.

Parallel to this runs the telemedicine layer with the hub-and-spoke network where doctors stationed at hubs consult patients located at peripheral centres.

On paper, the structure is clear. But mapping it revealed something more interesting: The system does not move in straight lines. It moves through people.

Medicines do not always arrive on time. Diagnostics are often unavailable locally. Private doctors are difficult to onboard due to cost structures. Hybrid practitioners fill gaps. Local and traditional healers remain accessible and trusted because they are present.

So patients do not simply “move up the referral chain” - they navigate availability. Simultaneously, telemedicine does not replace the physical system but overlays it and depends on it. Due to this interconnected structure, it is necessary to question what kind of a physical health system must exist for telemedicine to truly work.

If you’ve worked within layered public systems, especially in rural or resource-constrained contexts, we would love to hear your perspective. You can write to us at [email protected]. Until next time!

On day 5, we left early in the morning to meet one of the most widely known Khasi health providers, the traditional healers. For many Khasi people, they are the first point of care, especially for bones and teeth. Fractures, body aches, persistent pains, even something like gastritis, there’s a quiet confidence in the way people speak about their ability to treat it all using traditional methods. He gave one of us a quick massage, which showcased his years of experience, expertise, and knowledge.

When we met him, he showed us a stack of certificates, some even issued by the state government, almost like a bridge between belief and formal validation. When asked about formal medicine, he said that he believes every medicine system has its value, not just his own. When he falls sick himself, he goes to a hospital without hesitation.

And yet, he holds on to one idea very firmly: medicine has to remain personal. No screens, no remote prescriptions. A doctor should see the patient in front of them; that’s non-negotiable for him.

The next day, we set out for Jongksha PHC in the East Khasi Hill region. We spoke to a staff nurse who works closely with the e-Sanjeevani system on the doctor’s side. With the medical officer often not physically present, the staff nurse takes calls on his behalf, understands the case, and relays it to the doctor over the phone, then inputs the prescription into the system. 

For him, e-Sanjeevani works best as a record-keeping tool rather than a communication platform, especially since specialists are usually contacted directly to save time. He has also built small workarounds, like recording calls to avoid miscommunication during teleconsultations. 

While he supports increasing the teleconsultation target slightly, he’s careful about where to draw the line. Pushing it too high, he feels, would end up burdening the nurses at subcentres who are already stretched. For him, the idea behind increasing the target isn’t just numbers; it’s about slowly building awareness, letting patients ease into the system, especially those who don’t have easy access to doctors or specialists.

It’s less about scale, more about making the system feel usable in everyday life.

With the opportunity areas clearly defined, the process shifted from broad exploration to constructing more concrete concepts. Instead of generating more ideas, the focus now was on refining, combining, and selecting the most promising ideas or opportunities to move forward with from what had already emerged.

At this stage, three key questions guided the development of each concept:

What specific problem is this concept solving?

How does it work as a system, not just a feature?

What would make people adopt and trust this?

Structuring the Concepts

Rather than jumping to polished solutions, each concept was formulated by thinking through elements that would make it whole:

Core Idea – the central intervention.

Value Proposition – what are we offering to make the system better.

User Journey – how a vendor will navigate in the new system.

This helped develop a well thought out and structured concept rather than coming up with ideas and features.

Outcome of This Stage

The result of this phase is a set of early-stage concepts that are more cohesive than individual ideas but still open enough to evolve.

With a set of How Might We questions in place, we moved into concept generation. We began with a broad brainstorming session, intentionally keeping the space open to avoid locking into solutions too early.

This exercise helped us surface the overarching categories where most ideas naturally clustered. Across the board, ideas fell into a few recurring directions: awareness and assistance, instant or easier access to loans, repayment systems, and incentives to encourage continued engagement and repayment.

At this stage, we consciously stepped back to examine whether what we had generated were full ideas, individual features, or assumptions carried over from existing systems. This distinction became important—it helped us identify which directions had the potential to evolve into standalone concepts and which were better suited as supporting elements.

From there, we began refining and combining ideas, framing them into early concept directions rather than finished solutions. The focus remained on exploration: testing how different ideas might work together, what problems they addressed, and where they fell short.

Following fieldwork in both cities, we moved into a phase of synthesis and analysis. We revisited interviews with street vendors, bank officials, and urban local body (ULB) representatives, pulling out key quotes, recurring themes, and moments of friction.

This process helped surface not just what was said, but also what remained unsaid - patterns of dependency, trust, confusion, and workarounds that shaped how the scheme functioned on the ground.

This process helped us move into deriving insights by connecting patterns across interviews and contexts. These insights became the foundation for reframing the problem and guiding our next design decisions. From these insights, we moved into framing How Might We questions that helped us reframe the problem space and open up opportunities for exploration.