Computer Vision, OCR and Document AI Developer
I build document and image pipelines that turn photos, scans and PDFs into structured data your system can use, and send the unclear ones to a person.
133 completed contracts, 106 five-star reviews and a 100% Job Success Score on Upwork. 71% of that work came from clients who hired me again.
Where these projects go wrong
OCR demos run on clean scans. Real inputs are skewed phone photos, stamps over text and fields that move between versions of a form. The work is in confidence scores and in what happens to the pages the model is unsure about.
What I build
Document extraction
Fields pulled from invoices, IDs and forms with Google Vision, Document AI or open models, mapped to your schema.
Confidence and review queues
Low-confidence fields go to a person instead of silently into the database.
Face and image models
Face search, emotion detection and image generation pipelines, including LoRA training on GPUs.
Model licence checks
Every model checked for commercial use before it reaches paying users.
Proof
OCR document intelligence
Google Vision and Document AI extraction for a scanning app, plus passport and licence OCR.
Facial emotion detection app
An Expo app with OpenCV and TensorFlow models and LoRA training pipelines on FastAPI.
FaceSearch
Face search, LoRA training and virtual try-on in one SaaS, with GPU jobs queued.
Stack
Python, OpenCV, TensorFlow, PyTorch, scikit-learn, Google Vision, Document AI, FastAPI, Celery, Docker
Tell me what is broken
Send the workflow, the system or the idea. You get a straight answer on whether it is worth building, and what it would take. If I am not the right person, I will say so.