Lewis Roca

Lewis Roca, an Am Law 200 firm with 230+ attorneys across nine offices, faced a high-stakes construction dispute where their client — a top U.S. construction company — sued a subcontractor for failure to perform. The matter produced over 600,000 documents totaling 800 GB of data, spanning technically demanding file types including CAD drawings, PST archives, and proprietary construction project files. A short discovery turnaround and a fixed client budget compounded the pressure. The firm's existing eDiscovery tools were ill-equipped for data at this scale and complexity, and shuttling case data between separate analytics platforms created workflow friction that threatened both timeline and budget compliance.

After evaluating multiple platforms, Lewis Roca Partner Robert Roos selected Casepoint's eDiscovery platform based on its CaseAssist Active Learning functionality — a technology-assisted review (TAR) engine built on machine learning and predictive analytics — and its documented experience with construction-specific file formats. Partner and eDiscovery specialist Caitlin McHugh led day-to-day implementation, working directly with Casepoint's dedicated Client Success Manager and TAR consultants who provided workflow recommendations and iterative training guidance throughout the matter. This embedded support model allowed the team to continuously refine the active learning model, systematically surfacing relevant documents and eliminating non-relevant ones without the data-transfer overhead of their prior multi-tool stack. The unified platform removed the need to move data between systems, consolidating ingestion, analytics, and review into a single workflow.

Casepoint's AI-driven TAR reduced the reviewable document set by over 90%, collapsing what would have been an exhaustive linear review of 600,000+ documents into a fraction of that volume. The impact was measurable across both time and cost:

• >90% reduction in documents requiring attorney review
• 50% under budget — Lewis Roca completed discovery services well below the client's cost forecast
• 800 GB of complex construction data processed within a tight deadline

By automating the bulk of document triage, the legal team redirected their time from low-value review toward case strategy, improving both client value and matter profitability.

• Technology-assisted review (TAR/active learning) can reduce large document sets by 90%+, making fixed-budget discovery achievable even on matters with hundreds of thousands of records.
• File-type compatibility is a non-negotiable evaluation criterion: vendors lacking native support for CAD, PST, and construction project files create gaps that slow review and introduce risk.
• Embedded vendor expertise — dedicated project managers and TAR consultants — accelerates model training and improves precision; this support tier should be a procurement requirement, not a nice-to-have.
• A unified eDiscovery platform eliminates the friction and error risk of moving case data between analytics tools, materially improving throughput on time-sensitive matters.