Clarion AI

Inspiration

Medical lab reports are written for clinicians, not patients. Most people receive their blood test results and feel confused, anxious, or overwhelmed by unfamiliar medical terminology. We wanted to bridge that gap by building an AI system that translates complex lab data into clear, human-friendly explanations so patients can understand their health without needing a medical degree.

What it does

Clarion AI is an intelligent lab report interpreter. Users upload a PDF lab report, and the system automatically:

• Extracts text from the PDF
• Understands the lab values and medical meaning
• Explains results in plain English
• Highlights key findings and red flags
• Suggests questions patients can ask their doctor

It can detect patterns like anemia or abnormal white blood cell levels and explain what they mean in simple language so users instantly understand their results.

How we built it

Our pipeline combines document processing with AI reasoning:

Flow:
PDF → Text Extraction → AI Analysis → Explanation

Tech stack

• Next.js frontend for upload + UI
• PDF.js for document parsing
• Gemini AI for medical reasoning
• Node.js backend runtime
• Structured JSON schema for reliable output formatting

Challenges we ran into

• Parsing PDFs reliably across environments
• Handling scanned vs text-based reports
• Debugging worker issues with PDF libraries
• Fixing Windows permission and build cache errors
• Preventing AI hallucinations
• Making explanations medically accurate but simple

Accomplishments that we're proud of

• Built a full AI medical interpretation pipeline
• Successfully processed real lab reports
• Generated structured clinical explanations automatically
• Created a clean, beginner-friendly interface
• Ensured privacy (no medical data stored)

What we learned

• How to integrate AI with real-world health data
• Prompt engineering for structured medical output
• Debugging production-level runtime errors
• Designing UX for non-technical users
• Building reliable AI systems under time pressure

What's next for Clarion AI

We plan to expand Clarion AI into a full patient health interpreter platform with:

• Voice explanations for accessibility
• Multi-report trend analysis
• Risk prediction insights
• Wearable data integration
• Doctor-mode with clinical terminology
• Secure healthcare deployment

Long-term vision:
Make medical information understandable for everyone.

Built With

• ai
• api
• gemini
• html
• javascript
• next.js
• node.js
• pdf.js
• react
• typescript