How Relativity built AI for legal data intelligence with a foundation of trust on Azure

When you're handling the most confidential data imaginable—emails between a CEO and their General Counsel, sensitive government communications, evidence that can make or break a major litigation case—uncertainty about where that data goes or who has access to it isn't an option.

That's the reality for Relativity, a legal data intelligence company whose flagship SaaS platform, RelativityOne, helps the world's largest law firms, corporations and government agencies organize massive quantities of data to uncover the truth in legal matters. That data includes everything from emails and spreadsheets to WhatsApp messages and police body camera footage.

"The nature of legal data intelligence use cases, including internal investigations and litigation, involves highly sensitive data, messy data formats, and very tight timelines," says Kyle Disterheft, Senior Director of Product at Relativity who has been with the company for 14 years. “Adding to this, our customers have obligations to regulatory bodies and to their customers in how they handle the potential evidence and customer data used in these workflows.”

This creates a unique challenge: How do you bring cutting-edge AI capabilities to customers who have specific requirements around data privacy and security? For Relativity, the answer started with choosing the right partner.

A partnership built on trust

When Relativity began exploring how to integrate large language models into its customer solutions, it needed more than just powerful AI. It needed a partner with the capability to support strong safeguards around data processing, regional regulatory requirements, and security. With Azure, Relativity customers could be assured that once a data processing region was chosen by that customer for its instance, their data would remain in that region, helping to meet compliance requirements.

"Sometimes you have to make a compromise when you're choosing a vendor," says Aron Ahmadia, Vice President, Applied Science at Relativity. "You have to choose for scale. You have to choose for agility. You have to choose for trust. It's so rare that you have a partner that satisfies everything that you need and is constantly reinventing themselves to make a better offering. We're very fortunate to have such a great partner in Microsoft."

Microsoft could provide several must-haves for Relativity: never training models on customer data, respecting data processing requirements and providing enterprise-level services globally.

"There are two fundamental pieces of advice given to every attorney using AI for legal work in 2026," Ahmadia says. "Validate your work and know where your data is going. For any use of our product, the data processing question is the first question that's asked and it's also the last question that's asked."

These practices put firmly in place by Azure removed what Disterheft calls “immediate objections” in customer conversations, allowing teams to focus on the actual value AI could provide rather than getting stuck on trust concerns. Additionally, Azure’s global presence has allowed Relativity to deploy generative AI-powered products throughout the United States, the European Union, Canada, the United Kingdom, Australia, Japan, South Africa, Brazil, and more.

Transforming legal work at scale

Using Azure OpenAI models in Microsoft Foundry, Relativity aiR for Review can analyze documents at a pace that would be impossible for human reviewers alone. While a human attorney can review around 30 to 50 documents per hour, aiR for Review processes records hundreds of times faster with higher accuracy and greater consistency.

"Document review can be some of the most demanding, time-intensive work attorneys take on, and the resulting fatigue can increase the risk of human error and inconsistent decisions," Ahmadia says. "With aiR for Review, we’re empowering attorneys to gain greater and deeper understanding of the details in their documents during discovery, supporting them in surfacing insights more efficiently and ultimately making more confident, informed decisions."

More than 100 million legal decisions have been made with aiR for Review since launch, helping teams take on cases that were once impossible due to tight timelines and limited resources.

But it's more than just speed. Relativity aiR for Case Strategy extracts key facts from evidence and organizes them into clear timelines, summaries, and draft work product, empowering lawyers and litigation professionals to confidently construct the story of their case. That work would traditionally require junior associates or paralegals to manually comb through evidence for hours.

The partnership with Azure extends beyond AI models. Relativity relies on Azure Data Lake Storage as its foundational data service, handling everything from native document storage to the metadata that powers its core services. This data layer allows RelativityOne to handle the world’s largest and most complex matters, some with hundreds of millions of records.

By efficiently managing its data while scaling its usage, Relativity has reduced its Azure costs by 73% over a five-year period, all while consistently increasing throughput and import speeds.

Building for the future

Relativity continues exploring what else Azure can enable. The company is particularly interested in Azure OpenAI GPT-5's reasoning capabilities and efficiency, which could open up new use cases in legal data intelligence, especially as Relativity builds out further agentic AI capabilities across its solutions.

For Max Eggers, an engineering manager who's been with Relativity for nearly 10 years, that drive to explore what’s next defines the company.

"We've never rested on our laurels," he says. "We've always been challenging the status quo, trying to transform and evolve to meet the need or excel."

That relentless evolution has created something powerful: a platform where legal professionals can review documents exponentially faster than humanly possible, take on additional work they might have previously turned down because of the volume of data to review, and enable attorneys to spend their time on the high-value, strategic work that their clients hired them to do. All while keeping the most confidential data in the world exactly where it needs to stay. Through trusted AI and disciplined data controls, Relativity continues to advance the company vision: to elevate legal work and power the pursuit of justice.

Discover more about Relativity on Facebook, Instagram, LinkedIn, and YouTube.