Our Insights

Thought leadership and industry trends.

Home 9 Insights 9 AI 9 Beyond Keywords: Using Relativity aiR Custom Analyses to Extract the Details That Matter

Beyond Keywords: Using Relativity aiR Custom Analyses to Extract the Details That Matter

Nov 13, 2026

Legal teams often need to identify or extract specific information that metadata, keyword searches, dtSearch, or regular expressions alone cannot reliably pinpoint. In 5 in 5 Episode 19, CDS Senior Consultant Michelle Broxman and I explore how Relativity aiR custom analyses bridges the gap between merely finding words and effectively extracting their meaning.  

Relativity aiR Custom Analyses enables legal teams to define and deploy AI-driven review analyses using custom prompts tailored to their specific needs. By analyzing both searchable and non-searchable documents, custom analyses helps legal teams accomplish tasks that conventional search methods cannot reliably handle on their own.  

Beyond Relevance Categorization: Real-World Examples 

Custom Analyses goes well beyond categorizing documents as relevant or not relevant. It allows legal teams to define, deploy and scale document review analyses to fit the specific needs of any matter. The following examples illustrate how CDS has applied this capability to real-world challenges: 

Finding tractor photos while protecting privacy 

In a recent construction litigation matter, the CDS team collected data from a custodian’s phone and needed to identify photos containing a tractor without unnecessarily viewing personal photos. Using Custom Analyses, the team identified a narrow category of relevant visual information while limiting exposure to unrelated personal content.  

The payoff: Custom Analyses helped the team focus on relevant evidence while minimizing unnecessary exposure to personal information. 

Extracting PII and PHI from medical records 

In another matter, the CDS team needed to identify, extract, and redact personally identifiable information (PII) and personal health information (PHI) from patient medical records. Using Custom Analyses, the team created a prompt specifying the required data points. The analysis distinguished specific information, including birth dates without the year, and parent or guardian information.  

The payoff: Instead of indiscriminately extracting every date or address, Custom Analyses identified information that matched the requested categories. 

Applying context to numbers and time entries 

In a matter involving employee timesheets, the CDS team needed to extract check-in and check-out dates and times, calculate shift lengths, and identify shifts that resulted in overtime. Using Custom Analyses, the team created a custom prompt covering these related tasks and quickly obtained the results.  

The payoff: Custom Analyses demonstrated how context-based analysis can help legal teams interpret numerical information often difficult to extract through traditional text search alone.  

Looking ahead: Matter-specific analysis 

Relativity aiR Custom Analyses represents a significant advancement in how consultants and clients approach document review. Rather than simply asking AI to do more, Custom Analyses allows legal teams to define their own prompts in plain language and apply them across entire document sets. As investigation, litigation, and incident response teams continue to explore its capabilities, Custom Analyses has the potential to enable increasingly sophisticated, varied, and matter-specific workflows. 

Want to explore how Relativity aiR Custom Analyses could support your next matter? Talk to the CDS team at . 

About the Author

Nicole Guyer

Nicole Guyer

As Client Director at CDS, Nicole advises attorneys and ediscovery practitioners on managing electronic data through the EDRM lifecycle. She is passionate about developing sophisticated workflows and takes a hands-on consultative approach. She draws on more than a decade of eDiscovery experience to advise clients on advanced analytics and litigation support technologies to efficiently investigate data, cull data volumes and implement cost-saving solutions. Nicole has also managed large-scale document review projects and assisted with privilege reviews and privilege log creation. She is a Relativity Certified Administrator.