AI Document Data Extraction
Budget / SalaryHourly project
TypeFreelance project
LocationRemote
Posted1 hour ago
I have a growing archive of PDFs, Word files, and spreadsheets that I can’t keep up with manually. I want an AI workflow that reads each file, identifies the specific data points I care about, and writes them into a clean, structured table I can drop straight into Excel or a database.
The focus is purely analytical: no content generation, no customer-service chatbots—just reliable extraction. Accuracy matters more than speed, so I’m open to suggestions on the best mix of language models, OCR, or rule-based helpers to reach near-perfect capture rates across all three formats.
Deliverables I need from you:
• A working script or small app (Python preferred, but I’m flexible) that ingests mixed batches of PDFs, DOCX, and XLS(X) files and outputs a single, consistently structured CSV or JSON table.
• Clear documentation explaining how to add new fields or retrain the model if the document layout changes.
• A short test run on a set of sample files I will supply so we can confirm every required field is captured correctly.
If you have prior experience combining tools like PyPDF2, python-docx, pandas, spaCy, or similar libraries with machine-learning techniques to pull structured data, you’ll be a great fit. Let’s discuss your approach and the timeline you’d need.
The focus is purely analytical: no content generation, no customer-service chatbots—just reliable extraction. Accuracy matters more than speed, so I’m open to suggestions on the best mix of language models, OCR, or rule-based helpers to reach near-perfect capture rates across all three formats.
Deliverables I need from you:
• A working script or small app (Python preferred, but I’m flexible) that ingests mixed batches of PDFs, DOCX, and XLS(X) files and outputs a single, consistently structured CSV or JSON table.
• Clear documentation explaining how to add new fields or retrain the model if the document layout changes.
• A short test run on a set of sample files I will supply so we can confirm every required field is captured correctly.
If you have prior experience combining tools like PyPDF2, python-docx, pandas, spaCy, or similar libraries with machine-learning techniques to pull structured data, you’ll be a great fit. Let’s discuss your approach and the timeline you’d need.
Apply on Freelancer →
Project sourced from Freelancer.com. Applications happen directly on the original platform — we never collect your data.