INSURANCECENTRE_IDon’tKnowILostAShoe

Our project consists of a the member side gamifies insurance to make coverage easier to understand. The advisor side serves as an AI-powered CRM and productivity platform.

Comment

Inspiration

Insurance agents in Malaysia spend more time chasing spreadsheets and writing the same WhatsApp messages over and over than actually helping clients. We watched advisors miss renewal windows, forget follow-ups, and have no idea which client needed them most on any given day. We wanted to build something that felt less like a CRM and more like a co-pilot — something an agent would actually want to open first thing in the morning.

What it does

Our project is a gamified insurance advisor platform designed to help insurance agents manage their daily workflow, client relationships, and performance tracking in one place. The platform consists of two connected sides — a member-facing site where clients can view their insurance plans and manage their profiles, and an agent portal where advisors run their entire day from a single dashboard.

On the member side, clients can browse insurance plan categories, view their coverage status represented by a Protection HP bar that decreases as their policies approach expiry, log milestones and life events, and interact with their insurance information in a more engaging way. Whenever a client updates their life stage, occupation, or personal milestones, those changes are automatically reflected on the agent side, creating a live feedback loop between client activity and advisor awareness.

On the agent side, the dashboard is powered by several AI tools. The Smart Morning Brief ranks clients by urgency each day, highlighting customers who are approaching renewal, overdue for follow-up, or have recently experienced important life events. Advisors can open a client card, review the generated insights, and send an AI-drafted WhatsApp message with a single tap after manually editing it. Nothing is ever sent automatically.

The AI Tools page provides several productivity features. The Smart Morning Brief generates a daily summary of priorities. The Follow-up Draft Generator produces three message variations with different tones for each client. The Client Snapshot translates coverage information into plain-English talking points. The Renewal Radar analyzes multiple risk signals and categorizes customers into Critical, Attention, Watch, and Stable groups. The Life Event Signal identifies customers who recently experienced events such as marriage, newborn children, career changes, or retirement windows, helping advisors provide timely support. All AI outputs are generated server-side and clearly labelled for advisor review before any action is taken.

Beyond daily productivity, the platform includes a gamified progression system where advisors earn XP by completing tasks such as contacting clients, sending follow-ups, and reviewing policies. XP contributes toward Lucky Pack rewards, creating a lightweight progression layer that encourages consistent engagement without replacing the importance of client relationships.

A live leaderboard ranks advisors based on policies closed, revenue generated, consistency, and achievements, giving each advisor visibility into their performance and encouraging healthy competition within the branch. The Customers page provides a complete overview of every client, including policy status, Protection HP, life stage, and suggested actions, allowing advisors to manage their portfolio efficiently from a single platform.

How we built it

We built a two-sided platform — a member-facing site and an agent portal — sharing one codebase. The agent dashboard, AI tools, missions, leaderboard, and customer management pages were all built from scratch. Client risk scoring and life event detection run entirely on local logic with no API dependency. AI-powered features like draft generation and gap analysis call Google Gemini server-side so the API key never touches the browser. The whole thing compiles to a static export and ships as an Android APK via Capacitor so agents can use it on mobile without needing a server running.

Challenges we ran into

Getting the AI outputs to feel advisor-ready rather than generic was harder than expected — early drafts from Gemini were too formal and too long, so we spent real time tuning prompts with tone guides, word limits, and strict disclaimer rules. Building a custom styled dropdown to replace the native browser select element cost us more time than it should have; the OS renders native option lists in its own dark styling regardless of CSS, so we had to rebuild the whole interaction from scratch. Aligning the sidebar divider and the page header divider to sit on the same horizontal line was one of those pixel-perfect details that sounds trivial until you are forty-five minutes into it.

Accomplishments that we're proud of

The Morning Brief pagination and the WhatsApp draft sheet are the two moments where the product feels genuinely useful rather than just a demo. An agent can open the app, see exactly three prioritised clients, tap one, read a pre-written message that references the client's actual situation, edit it in seconds, and open WhatsApp with the message already filled in. That full flow — from zero context to ready-to-send — takes under a minute. We are also proud of the Life Event Signal loop: a member updates their profile on their side, and that change silently queues up as an actionable card on the agent's radar. No manual data entry, no lag.

What we learned

Gamification works best when the rewards feel earned rather than arbitrary. Early versions of the missions system felt hollow because the tasks were too abstract. Once we tied each mission directly to something an advisor would actually do in their real workflow — reviewing a lapsed client, running a protection gap check — the XP started to feel meaningful. We also learned that hiding complexity behind a clean interface is the real design challenge. The scoring engine behind Renewal Radar evaluates ten signals per client and does it instantly, but from the agent's perspective it just shows a coloured dot and a tier label. Getting that balance right between powerful under the hood and simple on the surface took most of our iteration time.

What's next for INSURANCECENTRE_IDon’tKnowILostAShoe

Moving forward, we envision INSURANCECENTRE_IDon’tKnowILostAShoe evolving into a more intelligent AI co-pilot for insurance advisors. Future enhancements include AI Project Memoria, which automatically records important customer milestones and WhatsApp conversations, a Life Event Prediction Engine that anticipates customer needs before they arise, an AI Schedule Coordinator that intelligently manages appointments based on advisor availability, and a Voice Memory Engine that transforms calls and voice notes into searchable knowledge. Together, these enhancements aim to reduce administrative work, improve customer relationships, and enable advisors to provide more timely and personalized services.

Built With

Share this project:

Updates