Last updated 08-22-2026
Category:
Reviews:
Join thousands of AI enthusiasts in the World of AI!
Referent
Referent lets solo lawyers and small firms run intake, email filing, deadline tracking, billing prep, and follow-ups through AI agents tied to live matter data. It pairs a legal CRM with an execution layer where agents stage work and lawyers approve every client-facing step before it goes out.
Most legal software is a system of record: it stores clients and matters and waits for you to type the next update. Referent positions itself as a system of action. Its agents open matters from intake links, file Gmail threads to the right case, draft replies for review, and assemble invoices while you practice law. That split matters if you want automation without handing judgment to a chatbot.
The platform connects Gmail, Google Calendar, and Google Drive today, with Outlook and Microsoft 365 marked as coming soon. It also runs an MCP server so ChatGPT, Claude, Perplexity, Gemini, and other MCP apps can query clients and matters with the same approval rules as the web app. iPhone and Android apps let you review agent work between hearings.
Referent is in a private beta with personal invites and white-glove onboarding. The team reports 450+ law firm applications for its first 20-seat cohort, closed August 7, 2026. Data stays in per-firm encrypted workspaces, with AES-256 at rest, TLS 1.2+ in transit, and no training on client data.
AI intake link screens leads, opens matters, and preps engagement letters for lawyer approval
Gmail integration files client emails and attachments to the correct matter automatically
Deadline agent tracks hearings, filings, and limitation periods with two-way Google Calendar sync
Billing prep assembles hourly or flat-fee invoices from captured time without manual timesheets
MCP server connects ChatGPT, Claude, Perplexity, Gemini, and OpenClaw to live firm data
iOS and Android apps on the App Store and Google Play for approvals between hearings
Agents act on live matter context instead of generic chat prompts
Lawyer approval gate on every client-facing or high-risk action
Combines CRM, automation, MCP access, and mobile apps in one workspace
Per-firm encrypted workspaces with no training on client data
White-glove onboarding migrates existing clients, matters, and files
Private beta access requires invite or application, not open self-serve pricing
Outlook and Microsoft 365 integrations are not available yet
Big Law tier is marked as coming soon on the marketing site
What is Referent?
Referent is AI-native legal practice management software for law firms. It combines a legal CRM with AI agents that handle intake, email filing, deadlines, billing prep, and follow-ups from your live matter context while lawyers approve every client-facing action.
How do Referent AI agents differ from legal chatbots?
Referent agents work from your firm's live clients, matters, emails, and deadlines rather than blank prompts. They open matters, file documents, draft follow-ups, and stage results for lawyer approval instead of only answering questions you paste in.
Does Referent integrate with Google Workspace?
Yes. Referent connects Gmail to file correspondence to matters, Google Calendar to sync hearings and deadlines both ways, and Google Drive to make linked documents searchable inside cases. Outlook and Microsoft 365 integrations are listed as coming soon on the site.
Is Referent free to try?
Referent offers a free signup at app.referent.law and runs a private beta with personal invites and white-glove onboarding. The marketing site does not publish seat pricing; access is by application or invite.
Does Referent have a mobile app?
Yes. Referent has iPhone and Android apps on the App Store and Google Play. The mobile apps let lawyers review agent work, approve critical actions, and check matters between hearings with the same permissions as the web workspace.
How does Referent protect client data?
Referent stores each firm's data in an isolated encrypted workspace with AES-256 at rest and TLS 1.2+ in transit. It does not train AI models on client data, logs every agent action in an audit trail, and requires lawyer approval before client-facing steps execute.