POD Document Automation System
Tech Stack: AWS Textract, SageMaker, FastAPI, React, React Native
Summary: Automated trucking document processing and offline signature verification using OCR and a deep learning model.
Details:
This project automated the processing of key trucking documents such as invoices, bills of lading, and pay orders. It used AWS Textract to extract data and a deep learning model trained on SageMaker to verify offline signatures. A FastAPI-based backend served real-time inference requests, while React and React Native handled user interfaces. The system significantly reduced manual paperwork and improved authenticity verification in logistics workflows.
ERP Order Automation from Email
Tech Stack: Selenium, BeautifulSoup, SpaCy, MongoDB, FastAPI
Summary: Automated Gmail-based order extraction and ERP integration using NLP and a bot dashboard.
Details:
Automated the entire workflow of order processing from Gmail by integrating Selenium bots and NLP. The system read order-related emails, extracted structured order data using BeautifulSoup and SpaCy, stored it in MongoDB, and auto-filled ERP forms using another Selenium bot. A FastAPI dashboard was built to track bot activity. The use of Pub/Sub ensured real-time email parsing without constant polling.
Travel Insurance Recommendation System
Tech Stack: PySpark, Sentiment Analysis, Apache Airflow, Flask
Summary: Recommended travel insurance plans using global risk datasets, sentiment scores, and real-time analytics.
Details:
A smart AI-driven platform to help travelers choose optimal insurance plans by considering various risk factors. Data from sources like global terrorism indexes, medical facilities, and weather conditions were aggregated. Sentiment scores were computed using web scrapers and NLP. Apache Airflow managed scheduled updates, and Flask hosted the front-facing application. The tool ensured personalized and up-to-date recommendations for travelers.
Leave a Comment