InvoiceOCR V2
Python, Fastapi,yolo deep-translator, pdf2docx, docx2pdf, cv2 Octuber 2024 – february 2025
• Developed software to translate Dutch invoices (images and PDFs) into English.
• Integrated PDF tools for merging, splitting, and PDF-to-image conversion.
• Used text extraction tools for selectable text in PDFs.
• Applied YOLO and Tesseract OCR for non-selectable text processing.
• Leveraged DeepTranslator for accurate language translation.
• Implemented logging modules for efficient debugging and tracking.
• Trained the system on a dataset of 10,000 invoices, requiring over two weeks for optimal training.
PDF2PDF
Python, Flask, pytesseract, pdf2docx, docx2pdf July 2024 – September 2024
• Translated data from Dutch to English in PDF format.
• Maintained the original table formatting within the PDF..
Fisca-quest| Python, Flask, MongoDB
• Web scraping of job listings using BeautifulSoup and Selenium.
• Added a scheduler that can schedule a task to scrape job listings at a given time.
• Developed a dashboard using Flask, Angular, and MongoDB where admins can view, search, and filter job
listings.
Chat-bot | FastAPI, MongoDB, and Rasa
• Labeled text with intents and entities for more than 4,000 words.
• Developed the website using FastAPI and the Jinja template, featuring a chatbot that provides FAQs to clients.
• Developed the back-end UI using FastAPI, Jinja, and MongoDB, where admins can add NLU data, train
models, and modify the pipeline.
• Deployed the project using Docker-compose.
Invoice-OCR | FastAPI, YOLOv4
• Labeled around 5,000 images using software named LabelImg.
• Developed a web API and a web application using FastAPI that utilizes a retrained YOLO model to provide
results.
• Performed image segmentation using YOLOv4.
• Deployed the project using Docker.