| --- |
| title: PaperPilot |
| emoji: 📄 |
| colorFrom: blue |
| colorTo: purple |
| sdk: gradio |
| sdk_version: "6.16.0" |
| python_version: "3.10" |
| app_file: app.py |
| pinned: false |
| tags: |
| - track:backyard |
| - sponsor:openbmb |
| - sponsor:openai |
| - sponsor:nvidia |
| - sponsor:modal |
| - achievement:offbrand |
| - achievement:fieldnotes |
| --- |
| |
| # 📄 PaperPilot |
|
|
| AI-powered Scholarship & Form Assistant |
|
|
| **PaperPilot** helps students and applicants understand lengthy application forms instantly by extracting key information, checking eligibility, generating document checklists, and answering questions using AI. |
|
|
| --- |
|
|
| # 🚀 Problem Statement |
|
|
| Many students and applicants struggle with: |
|
|
| * Long and complex application forms |
| * Hidden eligibility requirements |
| * Missing deadlines |
| * Missing mandatory documents |
| * Confusing instructions |
|
|
| PaperPilot solves this by automatically extracting and explaining the most important information from uploaded documents. |
|
|
| --- |
|
|
| ## Solution |
|
|
| PaperPilot uses OCR and AI to analyze forms and provide structured, easy-to-understand information. |
|
|
| Users simply upload a PDF and PaperPilot automatically extracts: |
|
|
| * Form Name |
| * Deadlines |
| * Eligibility Rules |
| * Required Documents |
| * Contact Information |
| * Form Summary |
|
|
| It also provides an AI assistant for asking questions about the uploaded form. |
|
|
| --- |
|
|
| # ✨ Features |
|
|
| ### 📄 Smart PDF Processing |
|
|
| * Supports normal PDFs |
| * Supports scanned PDFs |
| * OCR-based text extraction |
|
|
| ### 🤖 AI-Powered Form Understanding |
|
|
| * Automatic form analysis |
| * Structured information extraction |
| * Form summarization |
|
|
| ### ✅ Eligibility Analyzer |
|
|
| * Detects eligibility criteria |
| * Extracts income limits |
| * Identifies category requirements |
| * Highlights important conditions |
|
|
| ### 📋 Document Checklist Generator |
|
|
| * Extracts required documents |
| * Generates actionable checklists |
|
|
| ### 📅 Timeline Extraction |
|
|
| * Detects deadlines |
| * Highlights important dates |
|
|
| ### ⚠ Risk Detection |
|
|
| * Identifies missing information |
| * Detects critical deadlines |
| * Warns users about possible issues |
|
|
| ### 🔍 Document Verification Assistant |
|
|
| * Helps users verify required documents before submission |
|
|
| ### 💬 Ask PaperPilot |
|
|
| Powered by **Qwen 2.5 Instruct LLM** |
|
|
| Users can ask natural-language questions such as: |
|
|
| * What is the last date to apply? |
| * Am I eligible? |
| * What documents are required? |
| * Summarize this form. |
| * What happens if I miss the deadline? |
|
|
| --- |
|
|
| # 🏗 System Architecture |
|
|
| ```text |
| User Uploads PDF |
| │ |
| ▼ |
| OCR Engine |
| (Normal + Scanned PDFs) |
| │ |
| ▼ |
| Text Extraction |
| │ |
| ▼ |
| Master JSON Builder |
| │ |
| ┌──────┼──────┐ |
| ▼ ▼ ▼ |
| Summary Eligibility Checklist |
| │ │ │ |
| ▼ ▼ ▼ |
| Timeline Risk Detection Verification |
| │ |
| ▼ |
| Qwen 2.5 AI Assistant |
| │ |
| ▼ |
| User-Friendly Insights |
| ``` |
|
|
| --- |
|
|
| # 🛠 Tech Stack |
|
|
| ### Frontend |
|
|
| * Gradio |
|
|
| ### Backend |
|
|
| * Python |
|
|
| ### OCR |
|
|
| * EasyOCR |
| * PyMuPDF |
| * PDFPlumber |
|
|
| ### AI / NLP |
|
|
| * Hugging Face Inference API |
| * Qwen 2.5 Instruct |
|
|
| ### Deployment |
|
|
| * Hugging Face Spaces |
|
|
| ### DevOps |
|
|
| * GitHub Actions |
| * CI/CD Pipeline |
|
|
| --- |
|
|
| # 🔄 CI/CD Pipeline |
|
|
| ```text |
| GitHub Repository |
| │ |
| ▼ |
| GitHub Actions |
| │ |
| ▼ |
| Hugging Face Spaces |
| │ |
| ▼ |
| Automatic Deployment |
| ``` |
|
|
| Every push to the main branch automatically deploys the latest version of PaperPilot. |
|
|
| --- |
|
|
| # 🎯 Use Cases |
|
|
| * Scholarship Applications |
| * Government Schemes |
| * Admission Forms |
| * Internship Applications |
| * Job Applications |
| * Grant Applications |
| * Registration Forms |
|
|
| --- |
|
|
| ## Demo Video |
|
|
| https://youtu.be/EalpBFBLPA0?si=UVeDujcsYHECKKSm |
|
|
| ## Social Media Post |
|
|
| https://x.com/KamranX07/status/2066492830475038881 |
|
|
| ## GitHub Repository |
|
|
| https://github.com/KamranX07/PaperPilot |
|
|
| --- |
|
|
| ## 📝 Field Notes |
|
|
| ### Problem |
| Students often struggle to understand scholarship and application forms due to complex eligibility rules, deadlines, and document requirements. |
|
|
| ### Solution |
| PaperPilot is an AI-powered assistant that analyzes forms, extracts important information, verifies eligibility, and answers user questions in natural language. |
|
|
| ### Technical Architecture |
| - OCR/Text Extraction |
| - Structured Form Parser |
| - Master JSON Generation |
| - Eligibility Verification Engine |
| - Qwen 2.5 7B Instruct via Hugging Face Inference |
| - Gradio Frontend |
|
|
| ### Challenges Faced |
| - Hugging Face legacy inference endpoints were deprecated. |
| - Migrated from the old API route to the new InferenceClient-based routing system. |
| - Resolved dependency conflicts between Gradio and Pydantic. |
| - Improved eligibility extraction logic for income limits. |
|
|
| ### Lessons Learned |
| - Small language models can solve practical real-world problems effectively. |
| - Structured JSON extraction greatly improves reliability. |
| - Good UI/UX significantly improves user adoption. |
|
|
| ### Future Enhancements |
|
|
| * Multi-language support |
| * Advanced eligibility reasoning |
| * Personalized recommendations |
| * Form autofill assistance |
| * Voice-based interaction |
| * Mini-RAG document memory |
| * Mobile application |
|
|
| --- |
|
|
| # 👨💻 Team |
|
|
| - Hugging Face Username: KamranX07 (Solo) |
|
|
| Built as an AI-powered document intelligence platform for simplifying form understanding and application workflows. |
|
|
| --- |
|
|
| # ⭐ PaperPilot |
|
|
| **Upload. Understand. Apply with Confidence.** |
| **Built for the Build Small Hackathon.** |
|
|
