Inspiration
Manual document screening can be time-consuming, repetitive, and inconsistent, especially when dealing with a large number of documents. We wanted to build an AI-powered solution that could analyze documents efficiently, identify potential issues, and support faster, smarter decisions.
What it does
IRIS DocuScreen is an AI-powered document screening system that extracts important information from uploaded documents, analyzes their content, detects potential inconsistencies or tampering, and provides an intelligent screening result. Workflow: Upload → Extract → Preprocess → AI Analysis → Validate → Generate Result
How we built it
Frontend: React.js Backend: Python & FastAPI AI/ML: NLP, computer vision and machine learning Document Processing: OCR, PDF processing and image analysis Database: PostgreSQL Development: Git, GitHub and VS Code The system combines document processing, AI-based analysis, and validation into a single workflow to automate the screening process.
Challenges we ran into
Handling different document formats and image qualities Extracting accurate information from scanned documents Detecting altered, inconsistent, or suspicious information Improving AI accuracy while maintaining reasonable processing speed Protecting sensitive information and maintaining data privacy Integrating multiple AI and document-processing components into one system
Accomplishments that we're proud of
Built a complete AI-powered document screening workflow Integrated OCR, image processing, NLP, and machine learning Developed a functional React + FastAPI architecture Automated multiple stages of the document screening process Created a scalable foundation that can be extended with more AI models and validation techniques
What we learned
Applying AI and NLP to real-world document screening Combining OCR, image processing, and machine learning Building an end-to-end AI-powered web application Connecting AI models with a FastAPI backend and React frontend Handling unstructured and image-based document data Understanding the importance of accuracy, security, privacy, and reliability in AI systems
What's next for IRIS DocuScreen
We plan to make IRIS DocuScreen more accurate, robust, and scalable by improving tampering detection, expanding document-type support, strengthening AI-based validation, and adding more advanced detection capabilities. Our goal is to create a reliable platform that can significantly reduce manual effort and make document screening faster, smarter, and more secure.
Built With
- ai
- data-validation
- fastapi
- github
- image-processing
- javascript
- ml
- natural-language-processing
- ocr
- python
- react
