Casepoint puts the evidence in one place.

The construction company had plenty of lawyers. In Casepoint’s account, its panel numbered more than 300 outside counsel. What it lacked was a common view of discovery: the costly business of finding and reviewing potential evidence. Different firms used different systems. The client paid the bills, but comparing the work was difficult. The first thing to fail was visibility.

The bill nobody could compare

The company tested Casepoint across 27 matters involving 18 firms, 195 users and 14 terabytes of data. Monthly reporting made outside counsel’s discovery work easier to compare. According to the customer case study, discovery costs fell 57%; the client then required the platform on new matters. That is a vendor-published result from one customer, rather than a forecast for the next buyer. Still, the sequence matters: test, measure, then standardize.

The cloud was the early wager

Brothers Vishal and Vipul Rajpara founded Casepoint in 2008. The company’s origin story places its early wager on cloud discovery when legal organizations still relied on local systems. Vishal remains founder and chief technology officer; Vipul is now described as former chief operating officer. The business they built addresses an awkward institutional problem: the evidence belongs to the organization, while the process often spreads across its advisers.

The products follow that evidence through its working life. Legal Hold sends notices and reminders and tracks custodians. Cloud collection gathers relevant material. Processing prepares files for search; reviewers examine documents, apply redactions and produce material for a matter. Casepoint says its discovery software supports more than 600 file types. Filestore adds storage, while investigation tools help teams examine communications and relationships. The practical attraction is fewer transfers between stages.

Elastic offers a named example. Its legal team selected Casepoint through a competitive RFP, seeking stronger in-house discovery and legal-hold processes that worked with Google and Slack. In Casepoint’s published account, the team also used it to prepare contract-related emails for analysis in a separate AI research tool. A discovery platform had become useful for understanding ordinary business work. That extension is revealing: once information is accessible, another department’s question becomes cheaper to investigate.

The security review is part of the product

Casepoint sells into a crowded category. RelativityOne and Everlaw also offer cloud discovery and AI-assisted review. Calling software cloud-based no longer explains much. Casepoint’s particular position combines discovery with government workflows and demanding security environments. Its published security documentation lists FedRAMP High and Department of Defense Impact Level 5 and 6 authorizations. Those credentials matter to buyers handling sensitive information, although the applicable environment and authorization scope still need checking.

The distinction has a practical consequence. A team cannot choose a system on review speed alone if its security office cannot approve the deployment. Nor does a platform authorization make every customer workflow safe automatically. Permissions, collection boundaries and the treatment of privileged material remain decisions people must make. Casepoint’s expertise therefore includes implementation, migration and adoption support alongside the software itself.

One Fortune 100 migration account makes that work unusually concrete. Casepoint reports moving several thousand legal holds in 86 days. Two rounds of sample migration exposed seven mapping fixes; later testing found mismatches involving custodians who had been released and added again. The team corrected the migration logic before production. This is the unglamorous part of a platform switch: preserving the meaning of an old record, including its exceptions. A fast import is useful only if the receiving system still knows who is on hold.

A merger with a wider job description

In January 2025, Casepoint merged with OPEXUS alongside a majority investment from Thoma Bravo. Existing Casepoint shareholders and OPEXUS investor Gemspring retained minority positions. OPEXUS brought government process-management experience, including FOIA, audit and investigations. At the announcement, it reported more than 100,000 government users across 200 public institutions. The combination gave discovery technology a route into the surrounding administrative work. Financial terms were undisclosed.

The portfolio widened again with the July 2025 acquisition of mLINQS, a relocation-expense software company. Moving employees sounds distant from reviewing evidence, but both involve regulated processes and records that must survive scrutiny. Paul Colangelo became CEO in June 2026. On October 19, the company launched Casepoint Public Records for state, local and education organizations, connecting request intake with search, review, redaction and production.

The commercial model is tailored enterprise software. Current pricing guidance ties the price to applications and use cases. Buyers should specify their workload, storage, services and AI requirements before comparing quotes. The construction study’s per-gigabyte monthly arrangement is a historical customer example; it does not establish today’s rate for everyone.

Make the machine show its work

September 2026 brought Casepoint IQ, a shared framework for assistive AI, machine learning, predictive coding and agents. Its first two announced agents classify relevance and code documents against defined issues. The company says both were running with early-access customers. Users validate performance on a sample before wider execution, while decisions are recorded. The useful promise is inspectable automation: a reviewer can examine the reasoning and measure performance before committing a larger document population.

“They show their reasoning, measure their performance, and pause for human validation at critical points.”

Vishal Rajpara · on the IQ agents, September 2026

Copy the pilot, then count the handoffs

The repeatable lesson is a management method. Choose a bounded group of matters. Establish the baseline. Measure cost and turnaround. Check that migrated records match. Give outside counsel clear rules for access and reporting. Centralization is less compelling when an organization has little recurring discovery, cannot move required data, or lacks staff to govern the workflow. AI adds another condition: someone must define relevance and assess the sample. The software can organize the evidence; the institution still has to organize itself.