-
-
Whatsapp Interface
-
Vscode Interface
-
Github repo Interface
-
Main repo interface
Inspiration
We were inspired by a stark reality: for children with Thalassemia, a blood transfusion isn't an occasional medical procedure, it's their lifeline. We saw the immense emotional and logistical burden on patients and their families, who were forced to navigate clunky web portals or juggle phone calls in a desperate, real-time search for blood donors.
The inspiration wasn't to build another app that people would have to download and learn. It was the opposite. We asked: what if we could bring life-saving connections directly into the one app everyone already has on their phone? The solution had to be as simple as sending a text message, removing every possible barrier between a patient in need and a donor ready to help.
What it does
ThalaLink (BloodBridge Bot) is a smart assistant that lives on WhatsApp and SMS. It acts as a friendly, automated bridge between Thalassemia patients, blood donors, and healthcare providers.
For Patients
A patient can simply text "JOIN" to sign up. The bot then helps them manage their transfusion schedule, sends them helpful tips on nutrition and wellness, and provides daily emotional check-ins. If they report feeling unwell, it can instantly flag a human counselor for follow-up.
For Donors
A donor texts "I'M IN" and provides basic details. Our lightweight AI then predicts their next optimal donation date. When an urgent need arises nearby, the bot automatically notifies the best-matched donor, who can accept or decline with a single tap.
For Awareness
The bot turns every donor into an advocate. Through gamification, earning badges for donations and referrals, it encourages donors to share the bot with friends, creating a viral loop that organically spreads awareness and grows our network of heroes. In short, it automates the entire process of matching, scheduling, and support, all within a simple chat conversation.
How we built it
We designed ThalaLink with a "zero-install" philosophy, using a modern and scalable tech stack.
The Conversation Layer
We used Twilio to manage all communications on WhatsApp and SMS. This allows us to reach users on the platform they already trust.
The Bot's Brain
The conversational logic is powered by a Node.js application using Botpress, which allowed us to quickly design natural, multi-step dialogues for onboarding, matching, and support.
The AI-Powered Matchmaker
We built a separate microservice in Python using FastAPI. This service hosts a simple but effective XGBoost machine learning model that predicts donor availability, ensuring we always contact the right person at the right time.
The System's Memory
All user data, chat history, and gamification progress are stored securely in a PostgreSQL database.
Connecting the Dots
We integrated with external APIs like e-RaktKosh for donor directories and the Google Calendar API to automatically send schedule invites to both patient and donor once a match is confirmed.
Deployment
The entire system is containerized with Docker and deployed on Google Cloud Run, which allows us to scale effortlessly as our user base grows.
Challenges we ran into
Building this as a solo developer, far from the massive resources of Silicon Valley, meant every obstacle felt personal. The challenges weren't just technical; they were tests of resourcefulness and commitment.
The WhatsApp Gatekeeper
As a developer from Zimbabwe now living in South Africa, I know that WhatsApp isn't just an app; for many, it's the main channel for communication. The entire project's success depended on meeting people there. So, when I hit the notoriously slow WhatsApp Business API approval process, it felt like a roadblock. As a one-man team, I couldn't afford to simply wait for weeks. My solution was to build a fully functional SMS fallback from day one. It wasn't as elegant, but it ensured that the service would work for anyone, anywhere, right from the start, guaranteeing no one would be left behind while I navigated the bureaucracy.
The AI Cold-Start Problem
My vision was to use AI for intelligent matching, but AI needs data—a resource I had none of. I couldn't buy a dataset or wait months to collect information while people were struggling. The urgency of the problem was too great. So, I had to be pragmatic. I started by building a simpler, robust, rule-based matching system based on location and blood type. This allowed me to launch and begin providing immediate value. It wasn't the sophisticated system I dreamed of, but it worked, and every match it made helped build the very dataset I would need to power the AI in the future.
The Burden of Trust
Perhaps the biggest weight on my shoulders was data privacy. This wasn't a corporate project with a legal team; it was just me. I was asking a vulnerable community to trust a stranger—a single developer—with their sensitive health information. Earning and keeping that trust was non-negotiable. I spent countless hours implementing security best practices, ensuring every piece of data was encrypted and that consent was handled with the utmost clarity and respect. The responsibility was immense, because I knew a single mistake could shatter the trust the entire project was built on.
Accomplishments that we're proud of
We are incredibly proud of creating a solution that is profoundly human-centric. By eliminating the need for a new app, we've built something that respects users' time and avoids "app fatigue," making it more likely to be adopted and used.
We're also proud of the built-in viral growth mechanism. Instead of relying on a marketing budget, ThalaLink is designed to grow organically. Every successful donation and every referral from a happy donor helps the system reach more people, turning a simple tool into a self-sustaining community movement.
Finally, we're proud of the rapid development path. We proved that a high-impact, functional MVP could be built in just a few days, demonstrating the power of modern tools to solve critical real-world problems quickly.
What we learned
Meet Users Where They Are
The most powerful technology is often the one that feels invisible. The biggest lesson was that solving a communication problem doesn't always require a new interface; sometimes, the solution is to integrate seamlessly into existing behaviors and platforms like WhatsApp.
Build for Growth, Sustain for Impact
For social impact projects, sustainability is key. We learned that embedding growth mechanics like gamification and referrals directly into the product is more effective and sustainable than traditional marketing.
Pragmatism Over Perfection
The decision to start with a rule-based engine and evolve to an AI model was a crucial lesson in pragmatic development. It’s better to deliver 80% of the value now than to wait months to deliver 100%.
What's next for ThalaLink
Our roadmap is focused on a single goal: making ThalaLink the most reliable and efficient connection tool for the Thalassemia community. We will achieve this through a phased approach that prioritizes stability, user trust, and thoughtful growth.
Immediate Focus: Refine the Core Engine
Before adding new features, our priority is to perfect the existing ones. We will first deploy the smart matchmaking AI, transitioning from our initial rule-based system once we have sufficient data to ensure higher accuracy. Concurrently, we will strengthen our feedback loop by actively engaging our first users to identify and fix friction points. To support this, we will build a minimal, secure admin panel for our internal team to monitor system health and manage urgent escalations.
Next Steps: Drive Retention and Accessibility
Once the core service is stable, our focus will shift to making it stickier and available to more people. We will launch the gamification system, rolling out in-chat badges and referral streaks to make donating feel rewarding and turn donors into advocates. To ensure language is not a barrier to care, we will introduce support for key regional languages based on user demographics. We will also use data to optimize our notification system, refining the timing and content of reminders to reduce no-shows.
Long-Term Vision: Create a Sustainable Network
Our long-term goal is not to become an "everything app," but to build a self-sustaining network. We will deepen our partnerships with blood banks by providing them with anonymized, high-level data insights to help them optimize their operations. Instead of creating content, we will evolve the bot to act as a trusted signpost, connecting users to pre-vetted, localized support resources and certified counselors. Finally, to ensure no one is left behind, we will explore piloting a simple voice-call (IVR) based system for users with literacy challenges or limited smartphone access.
Log in or sign up for Devpost to join the conversation.