Klinical

Inspiration:

Imagine hearing that someone you love has a form of cancer that standard treatments can’t cure. You’re desperate for options, clinging to hope, but every search feels like hitting a wall of jargon, outdated trials, and endless uncertainty. That was our reality. What made it harder wasn’t just the diagnosis—it was the search for a lifeline. Weeks were spend digging through medical databases, comparing eligibility criteria, and juggling logistics, only to end up more confused and exhausted. We realized that hope shouldn’t be buried under complex systems. Klinical was born from that frustration and empathy—the idea that finding clinical trials should be as easy and human as asking a friend for help.

How we built it:

• Frontend: Lightweight HTML/CSS/JS interface for simplicity and responsiveness.
• Backend: Python (FastAPI) server handling user input and API calls.
• AI Integration: LLM (OpenAI GPT) to interpret unstructured text and summarize results.
• Data Source: Real-time clinical data from the ClinicalTrials.gov API.
• Architecture: Built around MCP-style design principles — connecting a model, contextual reasoning, and a live API tool into one smooth workflow.

What it does:

Klinical is an AI-driven clinical trial finder that transforms a single paragraph of natural language into real, personalized, easy-to-read trial results.

• Smart Text Understanding: Users simply describe their condition in their own words. Klinical’s AI interprets the text to extract key details such as age, condition, location, and treatment history. This eliminates the need for complex medical forms or terminology — just telling your story is enough.
• Real-Time Search: Klinical uses the ClinicalTrials.gov API to find verified, recruiting studies based on the extracted information. This ensures that all results are legitimate, current, and based on real clinical data.
• Plain-Language Summaries: The app rewrites dense medical descriptions into short, human-readable explanations and highlights key eligibility criteria. This makes clinical information approachable and inclusive for everyone.
• Simple Interface: Results appear as clean, organized trial cards with titles, locations, statuses, and links to official study pages. The goal is to keep the experience clear and focused so users can make informed decisions without feeling overwhelmed.

Challenges we ran into:

We started this hackathon working on something completely different — a hardware project using a Google Home Mini to manage medication reminders and responses. Unfortunately, we discovered that Google had retired conversational actions, meaning our concept was impossible to build by Saturday afternoon at 4 PM. It was a devastating moment, but we didn’t give up. Instead of quitting, we pivoted. Within a few hours, we brainstormed, researched, coded, and made Klinical — a project that not only worked but meant something to us personally. Turning that setback into something powerful and real is what we’re most proud of.

Accomplishments that we're proud of :

• We built Klinical from scratch in just a few hours after completely changing directions mid-hackathon. We took an idea inspired by personal experience and turned it into a working, meaningful product that could genuinely help people.
• We’re proud of our teamwork, creativity, and persistence — for not letting setbacks define us, and for staying up late, solving problems, and supporting each other until our idea came alive.

What we learned:

• Explore your tools early: We learned the importance of fully understanding technical limitations before diving in — like Google’s discontinued APIs.
• Natural ideas shine the brightest: The best projects come from something you truly care about, not something forced.
• Adaptability is everything: When one path closed, we found another — and it turned out even better.
• Leverage sponsor tools: Using various APIs and platforms from the hackathon sponsors guided us toward the right technical stack and helped us innovate faster.

What's next for Klinical:

• Transition to an MCP tool that can be used by other developers.
• Insurance and affordability insights: Help users understand potential costs or coverage for clinical trials.
• Worldwide accessibility: Expand beyond the U.S. to support international trial databases.
• Charity and funding options: Suggest organizations that help patients afford participation in research.
• Saved studies: Allow users to bookmark and track trials of interest for easier follow-up.
• Accessibility improvements: Add multilingual support and voice input for patients who struggle with typing or reading.

Built With

• api
• dedalus
• openai
• python