This year we celebrate the 13th annual Relativity Innovation Awards, and if there’s one theme running through every submission, I believe it’s this: the best AI in legal data intelligence is the kind that keeps a human in full control.
I’ve been in this industry for five and a half years now, and it’s all a blur. I still remember thinking “what even is e-discovery?” and honestly, I think my friends and family still don’t understand what I do for work. All I know is that I went from wrestling with the RDC (if you know, you know), to embracing the world of AI. I truly believe that, for the first time in a long time, there’s now an even playing field for everyone, which opens the door for so much creativity in our space.
At Relativity, spotlighting that creativity has always been one of our favorite traditions. But this year feels much different. We’ve watched generative AI move from skepticism to daily workflow. The temperament went from “we don’t really need AI,” to “we don’t trust it!” to “how do we use this on our terms and stay accountable for what it does?” The answers we’ve seen this year are really cool.
Back at Legalweek 2026, I overheard a line that stuck with me. Whoever you are, if you’re reading this ... it’s too late, I’ve already taken credit for it:
“If someone tells you they have the perfect AI, they’re lying to your face.”
I couldn’t agree more, because it’s exactly where we all are. Nobody has this fully figured out. We’re developing as AI is developing, learning in real time, refining as we go, and staying honest about what these tools can and can’t do yet. That’s what makes this year’s submissions worth celebrating. These solutions are genuinely impressive, and while these are awesome and finished applications, none of them claim to be perfect. They’re evidence of a community figuring it out together, out in the open, while keeping the final say firmly in human hands. And I can’t stress enough how important that is.
This year, the Innovation Awards submissions backed that up. We more than doubled the number of applications submitted for the Innovation categories. So, it’s very clear that our community isn’t just adopting AI; they're engineering everything around it to be accountable, efficient, and defensible.
Take a peek at our nominees for each technology-focused category this year.
Best Innovation: Organize
Data chaos is the enemy of data clarity, and these teams bring structure to the mess.
Bayer’s Bluefin automates the ingestion and routing of legal complaints, running more than 15,000 documents a year through a single system. Forensic Risk Alliance’s AIDE converts unstructured documents into structured, metadata-rich data, reporting up to 98 percent cost reduction in real deployments.
Meanwhile, HaystackID’s Relativity Migration Services has moved more than 25 terabytes into RelativityOne with no chain-of-custody failures, and Quinn Emanuel’s Microsoft 365 Metadata Bridge reassembles the parent-email and modern-attachment relationships that standard processing leaves broken.
Best Innovation: Discover
Discovery is a strategy, not just a phase, and you’ll see that these submissions sharpen how teams search, analyze, and interpret data.
EY Switzerland’s AI Deep Investigator proposes a structured investigation plan for the practitioner to review and refine before a single query runs, then tests competing hypotheses against the data. Deloitte’s Executive Summary Module turns a tagged corpus into a fact-based narrative where every claim is anchored to a Bates-numbered source and then runs an argumentative gap analysis on its own conclusions.
Epiq’s AI Accelerate brings OCR, translation, summarization, and image analysis together in one native RelativityOne interface, and Troutman Pepper Locke’s eMerge SearchCraft turns plain-English prompts into structured saved searches.
Best Innovation: Act
When the insight is in hand, someone still has to act on it – and carefully.
In support of that, EY Switzerland’s aiDetect & Redact (aiDR) leans on generative AI to auto-redact PII, PHI, and business secrets by jurisdiction, while reserving human judgment for the calls that truly need it. JND’s AutoRedact lets teams build and test detector profiles in a "Detector Lab" before running a full job, with confidence scores and QC baked in.
We’ve also seen how Lineal’s Amplify Productions ends the DAT-file troubles with a rules-based metadata engine that previews and validates every change before it runs. And Sevrak’s Relay allows you to describe a task in plain language, for which Relay drafts a plan showing every operation before it touches anything. Of course, it then waits for your approval.
Best Innovation: Workflows
Sometimes the best innovations come from admins and power users solving real problems with what’s already in the box.
IntrepidX’s ClearSet pre-screens document populations before they even hit AI analytics, and it’s built entirely on native RelativityOne capabilities. HaystackID’s Case Insight cuts review populations by more than 90 percent, taking fact development from weeks to days.
The team at NYS Information Technology Services equipped a simple PowerApps form to RelativityOne’s Legal Hold APIs so clients can kick off preservation themselves. And CDS’s DACH Intelligent Investigations tailors the aiR suite to the specific legal needs of the DACH region.
Best Innovation: Beyond
Lastly, some of the most interesting work this year lives “beyond” the edges of traditional e-discovery – addressing areas of legal data intelligence that are increasingly familiar to legal professionals across teams.
For example, Cognicion’s FieldBreaker wrangled a messy structured-data breach into a defensible notification report in 23 days. HaystackID’s VALID Suite gives legal teams a framework to authenticate digital media and flag AI-generated or manipulated content.
At PwC, their Intelligent eComms Monitoring puts aiR for Review to work surfacing compliance risk in employee communications that keyword search tends to miss. And MCS Group’s DispositionIQ takes RelativityOne all the way into physical records management from intake to destruction.
Who Will Take the Stage in Chicago?
Across all five categories, the unifying them is the same one I opened with: the best of this work keeps a human firmly in control. Nobody here claims to have AI figured out. They’re building in the open, honest about the limits, and leaving the final call to the people. That’s what makes this year’s class worth celebrating, and it’s why I’m excited for this year’s show and already looking forward to seeing what next year brings.
To explore every submission and vote for your favorites, visit the Community Choice ballot. Voting is open through August 22, and if you haven’t yet, there’s still time to register for Relativity Fest in Chicago (Sept 29–Oct 1), where we’ll celebrate this year’s winners.
