Inspiration
We’ve all seen how much time gets wasted on small but important tasks—whether it’s a farmer waiting for advice on crops, a student struggling for the right resources, or a shop owner stuck with paperwork. We asked ourselves: what if AI could quietly take care of these things in the background? That thought became our spark.
What it does Our project creates AI agents that don’t just talk, but actually do. They can automate a full workflow—cleaning and understanding data, making sense of it, and giving back real, actionable results. For example, a farmer could instantly know the best way to protect their crops, or a student could get personalized guidance without having to search endlessly.
How we built it We rolled up our sleeves and dove into IBM’s toolkit—Granite models for intelligence, preprocessing tools for data handling, and the Agent Development Kit to glue everything together. It wasn’t just writing code; it was experimenting, failing, fixing, and slowly shaping workflows that felt smooth and useful.
Challenges we ran into One big challenge was focus. There are so many areas in India where AI could help, but we had to choose a starting point. Another challenge was making sure the agents weren’t just “tech demos,” but something that feels natural and helpful to real people. That took lots of back-and-forth.
Accomplishments that we're proud of We learned that building AI isn’t just about fancy models—it’s about empathy. You need to understand what people actually need before you automate anything. We also discovered how powerful IBM’s ecosystem is when you start connecting the dots between its tools.
What we learned We learned that building AI isn’t just about fancy models—it’s about empathy. You need to understand what people actually need before you automate anything. We also discovered how powerful IBM’s ecosystem is when you start connecting the dots between its tools.
What's next for ZOE Zoe is still at the beginning of its journey. Right now, it’s a promising prototype, but we want to shape it into something people can actually use in their daily lives. Next, we’ll focus on making Zoe simple, friendly, and accessible—so a farmer, a student, or a small business owner can get help without needing to understand the tech behind it. Our dream is for Zoe to grow into a trusted companion that quietly takes care of tasks, while people focus on the things that truly matter to them.
Built With
- api
- docker-(for-containerization)-cloud-services:-ibm-cloud-for-deployment-and-scalability
- fhir-(fast-healthcare-interoperability-resources)-apis-for-health-data-exchange
- flask/fastapi-for-backend
- github-(version-control)
- hugging-face-(for-model-integration)
- javascript-frameworks:-ibm-agent-development-kit-(adk)
- kubernetes-for-orchestration-databases:-postgresql-for-structured-patient-records
- langchain-(workflow-orchestration)
- languages:-python
- mongodb-for-unstructured-healthcare-data-apis:-ibm-watson-apis-(for-nlp
- postman
- react.js-for-front-end-interfaces-platforms:-ibm-granite-open-source-foundation-models
- rest-apis-for-workflow-automation-other-tools:-pandas-&-numpy-(data-preprocessing)
- scikit-learn-(ml-utilities)
- speech-to-text)
Log in or sign up for Devpost to join the conversation.