Jen | Devpost

Inspiration

The inspiration behind Jen came from a deeply personal place—our grandma. Every time she needed a doctor’s appointment, she would rely on us to make the calls, check her health insurance details, and figure out her availability. It was a manual, often cumbersome process, but we always did it with love. One day, we thought, why can't there be an assistant that does all this for her? One that knows her needs, learns her preferences, and helps her in a personalized way? This dream to give our grandma more independence sparked the idea for Jen.

What It Does

Jen is an AI assistant designed to help elderly individuals with personalized support. It can book doctor appointments using the user's personal health information, insurance details, and availability preferences. When prompted to make an appointment, Jen contacts several hospitals, checks availability, cross-references insurance compatibility, and presents the user with suitable options. Once a choice is made, Jen takes care of confirming the booking. The idea is for Jen to be a companion that grows and adapts to the user’s needs, providing an evolving range of helpful services to make life simpler and more manageable.

How We Built It

We developed Jen using a diverse tech stack focused on performance and ease of use. Agentverse enabled multi-agent communication for complex tasks, while Fetch.ai facilitated decentralized real-time data exchange with hospital agents, restaurant agents, fun agents and etc. We used Groq for AI acceleration and OpenAI for natural language processing to ensure smooth, human-like interactions.

On the frontend, React and JavaScript created a simple, responsive interface. Python handled backend operations and third-party API integration, with VAPI allowing voice interaction. Perplexity powered Jen’s advanced search capabilities, and Deepgram was integrated for highly accurate voice recognition, enhancing accessibility for elderly users.

Challenges We Ran Into

Building Jen was far from easy. The most challenging part was integrating multiple disparate systems in a way that worked harmoniously. Going through documentation to connect different APIs—and finding that some were outdated or contained errors—was often exhausting. We faced latency issues, mismatches between documentation and actual API behavior, and even multiple conflicting documentations. However, these obstacles also taught us a lot about patience, troubleshooting, and maintaining our enthusiasm despite the setbacks.

Accomplishments That We're Proud Of

One of our biggest accomplishments is seeing Jen grow into a functional, helpful companion for the elderly. We’re proud of building an AI that is capable of understanding users, reaching out to multiple hospitals, dealing with complex data like insurance, and giving the user a clear and simple choice. Making a technology product that could make our grandma’s life easier is incredibly rewarding, and every successful interaction that Jen handles is a win for us.

What We Learned

Throughout this journey, we learned a lot about both technology and the needs of the elderly. We learned how to integrate empathy into technology, creating a product that listens and genuinely adapts to users. We also gained experience in managing the complexity of API integration, tackling unexpected challenges like incorrect documentation, and building a cohesive solution that elderly people could easily use. Most importantly, we learned about the real struggles faced by the elderly and how technology can help bridge that gap.

What's Next for Jen

Looking ahead, Jen has the potential to expand far beyond doctor appointments. We envision Jen handling prescription refills, arranging transportation, reminding users about events, and even helping with social activities. We want Jen to become a trusted companion—someone the elderly can depend on for both practical help and a sense of connection. Our mission is to empower the elderly, give them independence, and make sure they feel supported, no matter what challenges come their way.