Our read
The named examples split into two families. Business-law firms such as Hengeler Mueller, CMS, Goodwin and Reed Smith document Harvey or Legora for document-heavy work: due diligence, contract review, summaries, first drafts. Plaintiff firms such as Childers Schlueter, Batta Fulkerson and Geiger Legal Group document Supio, EvenUp or Eve for medical records and demand letters. Filevine appears only as an observed customer line with no AI task stated. These are examples to investigate, not a ranking.
Which tool for which job?
| Your job | Documented example | The reference to read | What it documents |
|---|---|---|---|
| First analysis of new laws and translation |  Harvey | Mori Hamada | Mori Hamada & Matsumoto lawyers use Harvey for a first analysis of new laws across jurisdictions, to draft a client update, and to translate and summarize foreign-language documents. 5 |
| Due diligence on large data rooms | Harvey |  Hengeler Mueller | Hengeler Mueller deal teams use Harvey Vault to work through large data rooms; the firm's story reports up to 95% less review time (vendor-reported). 2 |
| Contract review and issue lists | Harvey |  CMS | CMS lawyers use Harvey to analyze agreements and produce structured issue lists, and to review larger document sets across matters. 6 |
| Summaries, issue spotting and discovery drafting |  Legora |  Goodwin | Goodwin lawyers use Legora for summarization, issue spotting, term extraction and first drafts; litigators use it for document analysis and discovery drafting. 7 |
| Firm-wide rollout with certification | Harvey |  Reed Smith | Reed Smith rolled Harvey out across the firm from the start, without a small pilot, and set 100% AI certification for associates. 8 |
| Medical record questions in mass tort cases |  Supio |  Childers Schlueter | Childers Schlueter attorneys ask Supio questions about exposure, treatment and causation in Camp Lejeune cases and get answers linked to source records. 3 |
| Demand drafting and case assignment |  EvenUp |  Batta Fulkerson Law Group | Batta Fulkerson Law Group uses EvenUp to draft demands, flag missing bills and records, and review cases before assigning them to litigators. 4 |
| Demands, discovery and trial support |  Eve |  Geiger Legal Group | Geiger Legal Group uses Eve for demand letters, discovery work, direct-exam outlines and, during a trial, for counterarguments from medical records. 9 |
What work do law firms actually document with legal AI?
As of 2026-10-29, usedby's Legal AI category inside the Legal & Compliance sector lists 284 companies: 85 CONFIRMED by a customer story, quote or press release, and 199 OBSERVED on a vendor's own customer pages. Those counts show where public lines exist. They do not show quality, market share or return on investment.
Two families emerge. Harvey describes itself as a platform for legal research, drafting, contract analysis and due diligence, and Legora as a collaborative platform for document review, research and drafting. Supio, EvenUp and Eve are built for plaintiff work: Supio analyzes medical records and builds chronologies and demand letters, EvenUp covers the case lifecycle from intake to trial, and Eve adds intake qualification and nightly case reviews. Which family you need depends on whether your files are contracts and data rooms or medical records and claims.
Which document jobs do large firms give to Harvey and Legora?
Hengeler Mueller is the most concrete due diligence example: its deal teams use Harvey Vault to go through large data rooms, and the firm's customer story reports up to 95% less time on document review and up to 30% on summarization. Those are vendor-reported figures from one firm's deal context. The same story says EU-only data processing was part of the trust decision.
CMS documents three workflows with Harvey: contract review with structured issue lists, synthesis of curated legal materials, and review of larger document sets across jurisdictions. Mori Hamada & Matsumoto uses it for a first read of new laws and for translation that native speakers then check. Gleiss Lutz attorneys use it for first drafts, contract terms, closing checklists and company register excerpts. Goodwin documents Legora for summaries, issue spotting, term extraction and first drafts, plus discovery drafting and exhibit management in litigation.
These sources do not separate Harvey from Legora: both are shown on summaries, issue spotting and drafting. Our editorial suggestion, not a vendor claim: run the same set of your own contracts and one of your playbooks through both, inside Word. Harvey lists Word, iManage, NetDocuments and SharePoint among its integrations, and Legora offers a Word add-in with playbooks, so the document system you already use is the first criterion to check.
How did these firms get lawyers to use the tool?
Three rollouts, three shapes. Reed Smith skipped a small pilot and went firm-wide, with 100% associate certification; it reports 700 daily, 1,100 weekly and 1,400 monthly Harvey users among about 1,800 lawyers. Hengeler Mueller launched in six weeks with mandatory online training, in-person workshops in each office and an internal workflow competition, then kept a weekly short call and a shared library of use cases; the firm reports 96-97% adoption and around 4,000 prompts a day. CMS ran a twelve-month pilot with 300 lawyers on real matters before scaling.
Nothing here shows that broad beats pilot-first. What carries over to a smaller firm is cheap: a named person who coaches, a shared library of working prompts and workflows, and a certification expectation. The user counts and adoption rates describe firms with hundreds or thousands of lawyers and dedicated legal technology teams.
What do injury and mass tort firms use Supio, EvenUp and Eve for?
Childers Schlueter uses Supio on Camp Lejeune cases: attorneys ask plain-language questions about exposure, treatment and causation and receive answers with links to the source documents. Its partner says the system finds items in medical records the team did not know to look for, and its managing paralegal describes less burnout from record review. No time or settlement figure appears in the material read for this firm.
Batta Fulkerson Law Group, a 40-person firm with about 1,500 active injury cases, had one demand writer and a 45-day backlog. It uses EvenUp's full Demands and the faster AI-generated Express Demands for simpler cases, plus a case assistant for summaries and a view of missing bills and treatment gaps. The firm reports a month-long demand now taking about ten minutes, case assignment review cut by 75%, and more than 300 demands sent in four months. These are the firm's own reported results in a vendor case study.
Geiger Legal Group, described as a small firm, runs Litify and syncs records into Eve nightly. It reports demand letters ready for review in ten minutes instead of a day, roughly half its discovery workload handled by Eve, and counterarguments surfaced from medical records within ten minutes when the defense raised a new theory at trial. Mike Morse Law Firm is a CONFIRMED Eve customer whose public lines concern leadership: a Chief AI Officer and the fear of AI at law firms, not a stated task.
The three tools cover the same visible jobs, medical record review and demand drafting, and these sources do not tell them apart. Test them on one closed file from your own practice. Check the case-management link first: Supio lists Litify, SmartAdvocate, Casepeer, MyCase, Smokeball and Jove; EvenUp lists Litify, SmartAdvocate, CASEpeer, SharePoint, Google Drive and Dropbox. Check data location too: Eve's data processing terms state that customer data sits in US AWS regions at the time of writing.
Where does Filevine fit?
Filevine is case and matter management with LOIS, AI agents that draft in Word, research and build medical chronologies. In this sector it has OBSERVED lines only, such as Kopka Law Group and Singleton Schreiber. The Kopka line speaks of a growing firm needing more than a new case management system, which is a platform need, not a documented AI task. A data processing agreement is available and the vendor lists a free trial or free plan; no price is verified.
The criterion to settle is architectural: whether you want AI inside the system that holds your matters, or a separate layer that syncs from it, as Geiger Legal Group does between Litify and Eve. The sources here do not decide that for you.
What do the US peers in the record show?
Five US firms have CONFIRMED lines: Reed Smith (Harvey, firm-wide rollout), Goodwin (Legora, document and litigation tasks), Childers Schlueter (Supio, medical record review), Batta Fulkerson (EvenUp, demands and case review) and Mike Morse Law Firm (Eve, leadership quotes only). Three are OBSERVED, meaning the vendor shows them as customers: Kopka Law Group (Filevine), CunninghamLegal (Spellbook; the vendor's story headline concerns time saved in estate planning) and Akin (LegalOn). For those three, treat the record as adoption only; no task is established.
What can a smaller firm reuse, and what stays unknown?
The most portable examples are the plaintiff ones, because Batta Fulkerson and Geiger Legal Group are modest in size and describe a bottleneck most injury firms recognize: a demand backlog and slow file review. Before reusing either workflow, check your file mix, your case-management integration, where data is stored and who reviews each draft. Effort, scope and outcomes in these stories are the firms' own and do not transfer by default. No independent test of any tool appears in the record, and pricing is unknown for every tool except Filevine's free trial or plan.
Method and limits
2026-10-29
Scope: 284 companies classified in Legal & Compliance on usedby with a public line on a Legal AI tool. Every count on this page covers this sector only, not the tools' whole customer base. Company size is not recorded and is not claimed.
This dated snapshot counts distinct companies in public customer evidence. CONFIRMED means a case study, named testimonial or press source; DECLARED means the company itself told usedby; OBSERVED means a vendor listing; SIGNAL means a job post. All four count, but their strength differs. Removed and unlevelled lines are excluded. Examples are a capped selection verified within 90 days. Vendor facts are sourced records checked within 60 days. Coverage is uneven and missing evidence is unknown. These figures measure neither market share nor product performance; capture dates are not adoption dates.
Spellbook
LegalOn
