OCR & Clean Event Contacts
Budget: ₹12,500 – ₹37,500 INR
I have high-resolution JPEG images of the 2025 Australia mortgage-broking event attendee lists—everything is printed and fully legible. There are roughly 550 records to extract. I need a polished contact file that includes:
• Full name
• Company
• Email (DNS/MX-validated)
• Phone normalised to E.164 plus the local AU format
• City / state when shown
• Source-image reference and any clarification notes
Accuracy is paramount, so combine reliable OCR (Tesseract, Google Vision, ABBYY—or any equivalent you trust) with manual verification to catch edge cases and duplicates. Please normalise phone numbers, flag conflicting or missing data, and de-duplicate both exact and fuzzy matches.
Milestones
• Within 24 hours of project start: first 20 fully processed rows for my sign-off
• Final hand-off: no later than five business days from kickoff
Final package
1. Master file in both .xlsx and .csv
2. Separate “flagged rows” file highlighting issues or uncertainties
3. Brief QA report outlining methods, validation steps, and overall error rate
If something in the images looks ambiguous, drop it into the notes column rather than guessing—I’d rather review a flagged item than have a silent error.
• Full name
• Company
• Email (DNS/MX-validated)
• Phone normalised to E.164 plus the local AU format
• City / state when shown
• Source-image reference and any clarification notes
Accuracy is paramount, so combine reliable OCR (Tesseract, Google Vision, ABBYY—or any equivalent you trust) with manual verification to catch edge cases and duplicates. Please normalise phone numbers, flag conflicting or missing data, and de-duplicate both exact and fuzzy matches.
Milestones
• Within 24 hours of project start: first 20 fully processed rows for my sign-off
• Final hand-off: no later than five business days from kickoff
Final package
1. Master file in both .xlsx and .csv
2. Separate “flagged rows” file highlighting issues or uncertainties
3. Brief QA report outlining methods, validation steps, and overall error rate
If something in the images looks ambiguous, drop it into the notes column rather than guessing—I’d rather review a flagged item than have a silent error.
Related categories:
Data Processing
Data Entry
Excel
Web Scraping
OCR
Data Cleansing
Data Extraction
Data Management