From Discovery to Action: Building Raboti to Unlock Benefits for Reservist and Veteran Families

In this edition of our Innovation Chronicle, we invite you behind the scenes to share a pivotal moment for PollyLabs: when a discovery project became a live innovation effort. We introduce Raboti, our first in-house bot, which we believe has the potential to transform how families recover after disaster.

How it started

In February this year, we kicked off a discovery process focused on one central question:What are the real needs of reservists and their families, and where could technology meaningfully support them?

We chose to focus on this community for two reasons. First, it is one of the most critical populations in Israel today, carrying the sustained weight of long mobilization. Second, we were drawn to the chance to study a historically non-marginalized population suddenly hit by a crisis — a group we later called 'temporarily vulnerable, families who don’t typically rely on public support systems like social welfare, but who, after an unexpected life-altering event, face sharp, time-bound material challenges

Like all PollyLabs discovery work, we build a fact base using a mix of qualitative and quantitative methods, drawing on both primary and secondary data. In this case, we drew on insights from more than 500 reservists, 2,300 spouses, and 15 expert interviews. Through surveys, in-depth conversations, and request logs from a reservist's assistance center (request for assistance logs), we mapped urgent challenges across daily life, employment, mental health, and education.

Through this process, we surfaced six major areas of need that appeared consistently across geographies and demographics. These included educational regression among children, mental health strain, financial pressure, household stability challenges, and gaps in system navigation. Each theme reflected a different part of the burden families were carrying, and together they offered the first integrated view of the reservist family experience at scale.

The unexpected impact of discovery itself

Very early, it became clear that the discovery process had value beyond informing our next step. We documented and organized our key insights from the discovery effort in a report: Reserve Force Tech: Supporting reservist families: challenges, solutions, and the role of technology - and shared it with funders, civil society organizations, resilience centers, and government partners.

The response was immediate and widespread. Without any formal publicity effort, parts of the findings made it into national news. The document was forwarded through federations, foundations, and ministry teams. For many stakeholders, it was the first structured, bilingual overview of what families were experiencing and why they were struggling to access support. Months later, we continue to hear from organizations using the insights to shape programs, direct funding, and refine their own interventions. In other words, discovery did more than inform our work. It added a clarifying layer to those working in this relatively new field.

Identifying a leverage point

As we analyzed the six themes, we surfaced several areas where technology might have a meaningful role. To evaluate them, we used our tech repurposing framework (see visual) . 

PollyLabs Framework (read more here)

Benefits exist and families are eligible, but in practice the support is not reaching them. This gap appeared across interviews, regions, and family profiles. It is consequential, solvable, and deeply misaligned with the intention of existing policy.

This process led us to focus on rights utilization, the gap between what families are entitled to and what they are actually able to receive. Benefits exist and families are eligible, but in practice the support is not reaching them. This gap appeared across interviews, regions, and family profiles. It is consequential, solvable, and deeply misaligned with the intention of existing policy.

Why rights utilization matters?

As we went deeper, the scale of the problem of underutilization became impossible to ignore. Governments and private funders created a wide range of benefits for reservists and their families: financial support, educational support, tax adjustments, and other relief measures. Yet, across conversations and data points, families consistently described the same experience: the process of securing non-automatic benefits* felt hard to understand, fragmented across platforms, and overwhelming at a time when they were already exhausted.

This pattern is well documented in crisis recovery globally. Support often exists but remains unused, not because people do not need it, but because navigating complex systems requires time, attention, and emotional bandwidth that are in short supply during and after a crisis.

From a systems perspective, this creates a triple loss: resources underutilized, families unsupported, and institutions unable to deliver on their own commitments. In the long run, it creates a situation of risk creating a false narrative as if this is money that is not needed and can be reassigned and impact future allocation. 

From a PollyLabs perspective, it was exactly the kind of bottleneck where a clear, simple, and well-designed intervention could change outcomes.

Introducing Raboti

Built on the insights above and developed closely with our field partners, Raboti is our focused response to the rights utilization gap. It is a WhatsApp based chatbot designed to help Israeli reservist families understand and access the benefits already available to them.

Raboti - our new in-training WhatsApp bot

Raboti meets families in the channel they already use most - WhatsApp. They receive a link from a trusted partner, open a WhatsApp conversation that feels human and approachable, and are guided step by step through what they may be eligible for and how to apply.

Think of him as a Duolingo-style companion for your rights. Warm, direct, slightly funny, available around the clock, and supported by light behavioral nudges.

How did we build Raboti?

The first version of Raboti was simple. With our partners from the Reservists’ Wives Forum, we began by identifying which rights were most likely to be underutilized and mapped their eligibility and application steps. We then invited a group of twenty users to test an early journey using a Google Form, manual WhatsApp, and GPT guidance. As we refined the flow, we gradually expanded the group to fifty families.

Even in this fully manual stage, the model worked. Families used the prototype to apply for an average of about five hundred dollars within two days, with early indications of high unlocking and low rejection rates.

What will make his successful helping families?

Raboti is grounded in the principles that emerged directly from discovery and early prototyping: Meet people where they already are - WhatsApp is familiar, low effort, and widely used within this community. Make the experience feel human - Raboti communicates in a warm, direct, accessible tone. He is designed to feel like a companion, not an institution. Guide one step at a time - Families do not need long lists of benefits. They need clear, personalized direction. Keep it light - The MVP was built in weeks, not months, allowing us to test quickly and adjust based on real behavior before investing in additional infrastructure.

Where we go from here?

Raboti is now transitioning from being human-enabled to becoming a fully AI-powered bot. During this phase, we will test our hypotheses and evaluate where the user journey works well and where it needs adjustment. Our goal over the next four months is to serve one thousand families and unlock five hundred thousand dollars in existing benefits. Once achieved, we plan to scale to ten thousand families and then one hundred thousand, potentially expanding to other temporarily vulnerable groups, including displaced populations.

Once Raboti is stable in Israel, we will begin adapting the model for the US, with the first use case focused on transitioning veterans. They represent another temporarily vulnerable community facing similar challenges in navigating complex systems during periods of major disruption.

Temporary vulnerability is not unique to Israel or to military families. It appears after fires, floods, droughts, and other sudden disruptions. The systems differ, but the human experience is similar.

Raboti is the first in a family of chatbots designed to help people navigate essential systems during moments when clarity, simplicity, and support matter most.

A small, friendly guide for a very big challenge.

---

*marginalized non-automatic benefits - things you need to apply to and are not automatically given such as salary, bonus, etc.