Reconstructing Engineering Transmittal Histories Across Multiple Project Registers
Legal and engineering teams reconciled more than 22,000 engineering transmittals across multiple independent project registers — creating a defensible record of document movement for dispute review and evidentiary analysis.
Tens of thousands of transmittals. No single reliable record.
A major infrastructure programme had generated tens of thousands of engineering transmittals over several years of project delivery. The records existed across multiple independently maintained systems, each containing only part of the overall document history.
No single register could reliably confirm whether a transmittal had been issued, received, actioned, or omitted entirely. For legal and engineering teams involved in dispute review, reconstructing a defensible chronology of document movement required reconciliation across fragmented and inconsistent source records.
Reconciliation methodology, not just extraction
The core contribution was the design of the reconciliation methodology. Each source register used different identifiers, date formats, and status conventions. Aligning them required building a normalisation framework that could handle inconsistency at scale while preserving the relationships between records that mattered for the dispute.
AI-assisted extraction achieved 98.22% accuracy after calibration — enabling structured analysis across a volume of records that would have been impractical to process manually within the timelines of active litigation. But it was the cross-register comparison logic that made the findings defensible: identifying not just what each register contained, but what each one was missing.
From fragmented registers to a unified transmittal record
- Ingested transmittal records from multiple independently maintained source systems
- Extracted and structured key metadata fields using AI-assisted extraction, calibrated for accuracy
- Normalised identifiers, date formats, and status conventions across registers to enable direct comparison
- Matched corresponding entries across all three registers and identified missing, unmatched, and conflicting records
- Produced a unified Master Document List for review, analysis, and evidentiary use
What the reconciliation revealed
A significant number of transmittals appeared in only one or two source systems — confirming that no individual register represented the complete project history and that reliance on any single source would have produced an incomplete evidentiary record.
Analysis of correspondence series revealed missing entries and incomplete sequences across multiple project timelines — gaps that would not have been visible from any individual register and that required cross-register comparison to surface.
AI-assisted extraction and reconciliation enabled large-scale comparison of fragmented project records without full manual review — accelerating the identification of discrepancies and making the analysis tractable within active dispute timelines.
A defensible transmittal history the dispute could be built around
Legal and engineering teams received a unified Master Document List — a reconciled view of transmittal movement across all three source registers, with discrepancies, missing entries, and sequential gaps clearly identified and traceable to the underlying source records.
The structured reconciliation enabled teams to reconstruct document movement, identify conflicting transmittal records, analyse correspondence gaps across project timelines, and establish a defensible chronology for use in dispute review and evidentiary analysis.
In large infrastructure disputes, the evidentiary challenge is often not whether records exist, but whether the movement of those records can be reconstructed and defended across fragmented systems.