ClinIQ is an evidence-first medical research discovery and summarization platform. It bridges the gap between raw medical literature (PubMed) and actionable clinical insights by using advanced AI and deterministic ranking algorithms.
Note
AI-Enhanced, Not AI-Driven: ClinIQ prioritizes source reliability. All AI-generated summaries are strictly validated against a deterministic ranking engine to ensure ZERO hallucination of research papers.
- Intelligent Query Expansion: Automatically expands vague disease queries (e.g., "lung cancer") into comprehensive search terms covering treatments, clinical trials, and survival outcomes.
- Evidence Ranking Engine: A transparent scoring system that ranks papers based on publication recency, clinical intent, and relevance to the specific condition.
- Validated AI Summaries: Uses NVIDIA NIM (Llama 3.1 70B) to format complex research data into readable summaries.
- Hallucination Prevention: Features a built-in validation layer that rejects AI outputs if they mention papers or details not present in the original source data.
- Location-Aware Findings: Filters and prioritizes research relevant to specific geographic locations when provided.
- Framework: FastAPI (Python 3.12)
- AI Interface: NVIDIA NIM via AsyncOpenAI
- Data Source: PubMed E-Utils API
- Validation: Strict Pydantic-based data models
- Framework: React + Vite
- Styling: TailwindCSS
- Deployment: Vercel
- Python 3.12+
- Node.js 18+
- An NVIDIA NIM API Key
- Navigate to the
backenddirectory:cd backend - Create and activate a virtual environment:
python -m venv .venv source .venv/bin/activate # On Windows: .venv\Scripts\activate
- Install dependencies:
pip install -r requirements.txt
- Create a
.envfile:NVIDIA_NIM_API_KEY=your_key_here ALLOW_ORIGINS=http://localhost:5173
- Run the server:
uvicorn main:app --reload
- Navigate to the
frontenddirectory:cd frontend - Install dependencies:
npm install
- Create a
.envfile:VITE_API_URL=http://localhost:8000
- Run the development server:
npm run dev
- Set the root directory to
backend/. - Ensure
NVIDIA_NIM_API_KEYandALLOW_ORIGINS(your Vercel URL) are set in environment variables. - Railway will use the provided
railway.jsonandruntime.txtfor configuration.
- Set the root directory to
frontend/. - Add
VITE_API_URLpointing to your Railway service. - Ensure the Build Command is
npm run buildand Output Directory isdist.
Distributed under the MIT License. See LICENSE for more information.