Automate BOL/CI Data Extraction

₹600-1500 INR

Closed

Posted 7 days ago

₹600-1500 INR

Paid on delivery

I have roughly 5 000 PDFs that are a mix of Bills of Lading and Commercial Invoices. I need a reliable script—Python preferred, but any language is fine—that can open each file, read the key parties on the document, and aggregate everything into a single Excel workbook. The script must capture: • Shipper details • Receiver details • Broker details Accuracy matters more than speed; some files are machine-readable, others are scanned, so you may have to blend text parsing with OCR (think [login to view URL], PyPDF2, Camelot, Tesseract, or any stack you trust). The output should be a clean .xlsx file with one row per shipment and clearly labeled columns for each data point. Please send a brief but detailed proposal that explains: – The libraries or tools you will use – How you will handle both text and scanned PDFs – A plan for testing accuracy on a sample set before running the full batch – An estimated timeline for completion Deliver a runnable script, clear setup instructions, and the final Excel file. Let me know any questions you have so we can get started quickly.

Project ID: 40582233

About the project

10 proposals

Remote project

Active 3 days ago

Looking to make some money?

Email address

Benefits of bidding on Freelancer

Set your budget and timeframe

Get paid for your work

Outline your proposal

It's free to sign up and bid on jobs

10 freelancers are bidding on average ₹2,380 INR for this job

@muneeb6767

Hey, this is a great fit - I build this kind of PDF-to-Excel extraction pipeline regularly, including mixed batches of machine-readable and scanned docs like BOLs and commercial invoices. Approach: pdfplumber/PyPDF2 for machine-readable PDFs, Tesseract OCR (via pdf2image) for scanned ones. Layouts vary by shipper/broker, so I'll use keyword/positional anchors ("Shipper", "Consignee", "Broker") instead of rigid templates so it holds up across formats. Camelot if any fields need table extraction. Testing plan: before the full 5,000-file run, I'll process a sample of 30-50 files across your formats, verify shipper/receiver/broker fields against source PDFs, and refine logic based on mismatches. Once sample accuracy is solid, I run the full batch, with a status column flagging low-confidence files for review. Output: single .xlsx, one row per shipment, columns for Shipper, Receiver, and Broker (name/address/contact), plus a confidence/flag column. Deliverables: runnable Python script, short setup guide (dependencies, pointing at your PDF folder, rerunning on new batches), and the final Excel file. Timeline: sample + validation in the first couple days, then full run once logic's confirmed - roughly 5-7 days depending on how many scanned/low-quality files show up. Question: are shipper/receiver/broker fields labeled consistently across documents, or does it vary a lot by carrier/broker? Affects how much I lean on OCR + fuzzy matching vs straightforward parsing.

₹1,500 INR in 2 days

5.0

(47 reviews)

5.6

@Ahmadw208

Hi, Could you provide a sample of the PDFs so I can assess the complexity involved? To tackle your project, I’ll use Python with libraries like PyPDF2 for machine-readable PDFs and Tesseract for OCR on scanned files. My approach involves initially parsing documents, extracting shipper, receiver, and broker details, then aggregating everything into an organized Excel workbook. I will implement a testing phase on a small sample of PDFs to ensure accuracy before processing the entire batch. The complete project should take around 5-7 days, including testing and revisions. You’ll receive a runnable script with thorough setup instructions and the final .xlsx file. Looking forward to your response! Best Regards, Ahmad

₹1,050 INR in 7 days

0.0

(0 reviews)

0.0

@orwells

Hi, this is squarely what I do. I build Python pipelines that read messy PDF batches and turn them into one clean Excel sheet, so 5000 mixed Bills of Lading and Commercial Invoices into a single workbook is a job I can run reliably. My approach is text first, OCR fallback. I parse the machine readable files directly with pdfminer and Camelot for tables, and route the scanned ones through Tesseract, so both kinds get read accurately. For each file I pull shipper, receiver and broker details and write one row per shipment with clearly labelled columns. I add validation flags so any low confidence read is easy to spot and fix. Since you said accuracy beats speed, I will run a small sample first so you can confirm the columns before the full batch. Send me a few sample PDFs and I will start.

₹1,050 INR in 7 days