🏷️ Project Title

OCR & PDF Text Extraction Microservice

🧾 Executive Summary

The Personal AI Factory v1 – OCR & PDF Text Extraction Microservice is a production-grade, serverless backend component designed to extract clean, machine-readable text from textual PDF documents. Built using TypeScript and deployed as a Vercel Serverless Function, this microservice exposes a single HTTP API endpoint that fetches a remote PDF file, parses its textual content using pdf-parse, and returns structured JSON output.

This service is optimized for automation-first architectures, specifically downstream integration with n8n pipelines. Version 1 explicitly supports text-based PDFs only and does not perform OCR on scanned documents or images.

📑 Table of Contents

1. 🧩 Project Overview
2. 🎯 Objectives & Goals
3. ✅ Acceptance Criteria
4. 💻 Prerequisites
5. ⚙️ Installation & Setup
6. 🔗 API Documentation
7. 🖥️ UI / Frontend
8. 🔢 Status Codes
9. 🚀 Features
10. 🧱 Tech Stack & Architecture
11. 🛠️ Workflow & Implementation
12. 🧪 Testing & Validation
13. 🔍 Validation Summary
14. 🧰 Verification Testing Tools
15. 🧯 Troubleshooting & Debugging
16. 🔒 Security & Secrets
17. ☁️ Deployment (Vercel)
18. ⚡ Quick-Start Cheat Sheet
19. 🧾 Usage Notes
20. 🧠 Performance & Optimization
21. 🌟 Enhancements & Features
22. 🧩 Maintenance & Future Work
23. 🏆 Key Achievements
24. 🧮 High-Level Architecture
25. 🗂️ Folder Structure
26. 🧭 How to Demonstrate Live
27. 💡 Summary, Closure & Compliance

🧩 Project Overview

This microservice functions as a stateless PDF text extraction API within the Personal AI Factory ecosystem.

• Accepts a publicly accessible PDF URL
• Downloads the PDF at runtime
• Parses textual content using pdf-parse
• Returns extracted text as structured JSON
• Designed for synchronous HTTP execution

🎯 Objectives & Goals

• Provide a reliable text extraction layer for automation workflows
• Eliminate dependency on paid OCR services for textual PDFs
• Maintain fast cold-start and execution times
• Enable seamless integration with n8n HTTP Request nodes
• Serve as a foundational V1 component for future OCR and AI expansion

✅ Acceptance Criteria

• HTTP 200 returned for valid textual PDFs
• Structured error responses for invalid input
• No OCR processing in Version 1
• Deployable on Vercel without custom infrastructure
• JSON output compatible with automation tools

💻 Prerequisites

• Node.js 18 or higher
• Vercel CLI (for deployment)
• Publicly accessible PDF URLs
• Basic REST API knowledge

⚙️ Installation & Setup

1. Clone the repository
2. Install dependencies
3. Verify Node.js version compatibility
4. Review TypeScript configuration
5. Prepare Vercel deployment

🔗 API Documentation

Endpoint: /api/ocr-summarize

Method: GET / POST

Input: Public PDF URL (fileURL)

Output: JSON with extracted text

🖥️ UI / Frontend

This project does not include a frontend or UI layer. It is designed for backend-to-backend and automation-based consumption via n8n, Postman, or Curl.

🔢 Status Codes

Status	Description
200	Successful extraction
400	Invalid or missing fileURL
500	Internal server error

🚀 Features

• Textual PDF parsing
• Serverless execution
• Automation-friendly JSON output
• No paid OCR dependencies

🧱 Tech Stack & Architecture

• Runtime: Vercel Serverless Functions (Node.js 18)
• Language: TypeScript
• PDF Parsing: pdf-parse
• HTTP Client: node-fetch
• Deployment: Vercel

Client / n8n
     |
     v
Vercel Serverless Function
     |
     v
pdf-parse
     |
     v
JSON Response
  

🛠️ Workflow & Implementation

1. Receive HTTP request
2. Validate input parameters
3. Fetch PDF from URL
4. Parse text using pdf-parse
5. Return structured JSON response

🧪 Testing & Validation

ID	Area	Command	Expected Output	Explanation
T-01	API	GET with valid PDF	200 + text	Valid textual PDF
T-02	API	Missing fileURL	400 error	Validation check

🔍 Validation Summary

• Input validation enforced
• Error handling implemented
• Automation compatibility verified

🧰 Verification Testing Tools

• Curl
• Postman
• n8n HTTP Request node

🧯 Troubleshooting & Debugging

• Ensure PDF is publicly accessible
• Confirm PDF is text-based
• Check Vercel function logs

🔒 Security & Secrets

• No secrets or API keys required
• Stateless execution
• Public-file access only

☁️ Deployment (Vercel)

• Node.js 18 runtime
• 2048 MB memory
• 60-second execution limit

⚡ Quick-Start Cheat Sheet

1. Deploy to Vercel
2. Copy endpoint URL
3. Provide public PDF URL
4. Receive extracted text

🧾 Usage Notes

• Textual PDFs only
• No OCR support in V1
• Designed for preprocessing pipelines

🧠 Performance & Optimization

• Lightweight dependencies
• Fast cold-start execution
• Performance dependent on PDF size

🌟 Enhancements & Features

Current version supports textual PDF extraction only.

🧩 Maintenance & Future Work

• OCR for scanned PDFs
• AI summarization layer
• Chunking and vector storage

🏆 Key Achievements

• Production-ready serverless microservice
• Zero-cost alternative for textual PDF extraction
• Automation-first design

🧮 High-Level Architecture

The service acts as an independent extraction node within the Personal AI Factory, feeding structured text into downstream automation and AI systems.

🗂️ Folder Structure

ocr-summarizer-microservice/
├── api/
│   └── ocr-summarize.ts
├── types/
│   └── pdf-parse.d.ts
├── node_modules/
├── package.json
├── tsconfig.json
├── README.md
└── .gitignore
  

🧭 How to Demonstrate Live

1. Deploy to Vercel
2. Send GET request with PDF URL
3. Observe extracted text response

💡 Summary, Closure & Compliance

This repository delivers a compliant, production-ready, serverless PDF text extraction microservice aligned with Personal AI Factory v1 standards.

License: MIT
Author: Ansh Srivastava
Status: Stable – Production Ready (V1)