IntrepidX Cuts $46,800 in Privilege Review Costs and Uncovers Hundreds of Unknown Attorneys with Relativity aiR for Privilege

Customer Since
2021

Headquarters Location
San Diego, CA

Share Their Story

How did they do it?

  • Cut 500,000 documents to 67,000 responsive documents using aiR for Review
  • Ran the remaining documents through aiR for Privilege, generating 10,000 privilege log descriptions overnight
  • Uncovered 349 attorneys the case team had no prior knowledge of, ensuring privileged material stayed protected and out of production
  • Completed the full privilege workflow in under two weeks with a lean team – work that would have otherwise required 35 reviewers and 3 managers to achieve manually

Half a Million Documents. Two Weeks. No Room for Error.

IntrepidX is a full-service e-discovery partner known for managing complex, high-pressure litigation matters with precision and care. When a 500,000 post search term document matter came to them with a tight deadline, they got to work immediately using Relativity aiR. Through initial culling and first-pass review with aiR for Review, they brought the population down to 67,000 responsive documents, which also saved the client the expense of a large-scale review. Completing the remaining privilege review thoroughly, accurately, and on time was a challenge that a traditional approach wasn't built to handle.

Finding What Nobody Knew Was There

Going into the matter, the case team had identified 10 attorneys. When they ran aiR for Privilege across those 67,000 documents, the full picture looked very different: 359 attorneys across 12 law firms. The 349 names nobody had accounted for weren't a minor gap – each one represented documents that could have made it into production without ever being flagged. With the full attorney population identified, the team could be confident no privileged document would reach production unreviewed.

Beyond catching what was there, aiR also provided documented rationale for every decision, giving the team a clear, reviewable record for each call. For a client engaging with AI-assisted privilege review for the first time, that transparency was essential – reasoning they could follow, stand behind, and defend if challenged.

From 300 Hours of Manual Work to an Overnight Run

Privilege logs that would have taken IntrepidX hundreds of hours to write were ready overnight – 10,000 drafted descriptions waiting when the team returned the next morning.

"AI didn't replace reviewer expertise, it amplified it. With 10,000 privilege log descriptions generated overnight, our team focused on the nuanced privilege analysis and judgment calls that matter most, like when and where redactions were appropriate. Automating time consuming tasks allowed reviewers to spend time validating decisions, protecting privileged information, and delivering the quality and defensibility our clients expect."
Catherine Pray, Senior Director, Managed Review, IntrepidX

That kind of shift doesn't just reduce the burden on reviewers – it fundamentally changes what's achievable on a tight timeline. With description drafting handled, the team focused entirely on the review decisions that required human judgment.

A Smarter Path through Privilege Review

Using aiR for Privilege’s document predictions, IntrepidX built a QC process around risk rather than volume. Documents flagged as wholly privileged or requiring redaction received direct review; others were sampled based on the nature of the flag to determine the appropriate level of review depth.

The impact was significant. The targeted approach cut $46,800 in privilege review costs compared to full manual review of the same population, and the team wrapped the entire workflow in under two weeks – work that would have required 35 reviewers and three team leads to hit that same timeline manually. IntrepidX delivered the same quality and completeness with significantly fewer resources.

On a matter where completeness, speed, and defensibility were all essential, aiR for Privilege delivered on each one.

Ready to see what Relativity aiR can do for you?