We successfully migrated 15,000+ legacy contract documents from our old system into Adobe Experience Cloud AEC-2023, establishing a foundation for our new e-signature workflow. The project took 6 weeks and involved significant format standardization work.
Our legacy contracts came in mixed formats (PDF, Word, scanned images) with inconsistent naming conventions and metadata. We needed to standardize everything to enable the Document Cloud e-signature integration while preserving audit trails and contract relationships.
Key challenges we addressed: converting all documents to searchable PDFs, mapping legacy metadata fields to AEC’s contract management schema, and configuring automated routing rules for signature workflows. The Bulk Document Import tool handled the volume, but we had to build custom preprocessing scripts.
The result: contracts now flow automatically to Adobe Sign based on contract type and value thresholds. Approval times dropped from 8 days to 2 days average. Our legal team can track every signature event with full audit compliance.
What preprocessing scripts did you build for format standardization? We have similar mixed-format challenges - lots of scanned images that aren’t searchable. Did you use OCR processing before import?
We took a hybrid approach. Full historical access logs from the legacy system were exported to a separate audit archive database that compliance can query when needed. For AEC, we imported the complete document version history (where it existed) and created audit notes summarizing key legacy events - original creation date, major amendments, approval milestones.
Each migrated contract has a “Migration Audit Summary” attachment that lists legacy system events. This satisfied our auditors while keeping AEC’s database lean. The archive link is stored in a custom field so users can drill into detailed legacy history if required.
Yes, OCR was essential. We used Adobe’s Document Services API for batch OCR processing of scanned images and non-searchable PDFs. The preprocessing pipeline had three stages: format conversion (Word/Excel to PDF), OCR application for image-based files, and quality validation.
Quality validation was critical - we checked text extraction confidence scores and flagged low-quality scans for manual review. About 400 documents needed rescanning from physical originals. The entire preprocessing ran on cloud infrastructure to handle the volume efficiently, processing roughly 500 documents per hour.
We extended the schema strategically. AEC’s base contract object gave us 80% of what we needed - standard fields like contract number, parties, dates, values worked perfectly. We added 12 custom fields for our specific needs: legacy_system_id (critical for traceability), renewal_notice_days, auto_renew_flag, and department_code among others.
The mapping logic ran through a validation layer before import. Any record with missing required fields went to an exception queue for manual review. About 8% needed human intervention, mostly for contracts with ambiguous party names or missing effective dates.