mixus

mixus

An AI agent platform for law firms: assign work by email, get reviewable redlines and deliverables back in Word, approved by lawyers.

paid
Last Updated 8/31/2026

About mixus

mixus is an AI agent platform for law firms, designed around the email and Word workflows lawyers actually use. Users forward a matter to the agent by email and CC those who need to review; the agent completes the work in the same thread and returns reviewable .docx and .xlsx files, such as redlines, drafts, analyses, and spreadsheets. Lawyers review the changes, issues, and citations before approving. The platform emphasizes firm-level oversight: every redline is tied to a firm-defined playbook rule, and the system learns from reviewers' accept/reject decisions, while partners can view firm-wide analytics on playbook usage. On security, the official website states that its infrastructure is SOC 2 Type II certified and compliant with ISO 27001:2022, GDPR, and HIPAA; it uses TLS 1.3 for data in transit and AES-256 for data at rest, supports customer-managed keys (BYOK), and AI calls use Anthropic's zero-data-retention API, so prompts and outputs are not stored or used for training by the model provider. It also offers matter-level data isolation, exportable audit records of every edit, a zero-retention mode for regulated work, and US and EU data residency options. The official website states it is used by Am Law 10 firms.

Best for

Law firms handling confidential client matters with data security and compliance requirements
Lawyers and paralegals who do most of their work in email and Word
Law firm security and compliance teams that need to vet AI tools

Best use cases

Forward a matter email to the AI agent and CC reviewers, receiving redlines, drafts, analyses, or spreadsheets in the same thread

Lawyers review the AI-produced changes, issues, and citations before approving the work product

Security teams review SOC 2 Type II, encryption, key management, model provider data retention, subprocessor, and data residency materials before deployment

Compare before choosing

The official comparison page states: Claude is a chatbot, while mixus runs workflows — the same AI foundation, with mixus adding the playbooks firms actually use, accumulated experience, email handoffs, and organizational controls on top.

The official comparison page states: compared with Claude for Word, mixus differs in enforcement — every redline is tied to the firm's custom playbook rules, learns over time from reviewers' accept/reject decisions, and gives partners firm-wide analytics on playbook usage.

Features

Email-driven workflow

Forward a matter to the agent and CC reviewers; the agent completes the work in the same thread.

Reviewable work products

Redlines, drafts, analyses, and spreadsheets are returned as reviewable .docx and .xlsx files.

Lawyer review and approval

Lawyers review the changes, issues, and citations produced by the agent before approving.

Playbook rules and learning

Every redline is tied to a firm-defined playbook rule, and the system learns from reviewers' accept/reject decisions.

Firm-wide visibility

Partners can view firm-wide analytics on playbook usage, as well as usage, approvals, and outcomes across matters, reviewers, and playbooks.

Security and compliance

SOC 2 Type II, ISO 27001:2022, GDPR, and HIPAA; TLS 1.3 encryption in transit and AES-256 encryption at rest; customer-managed keys (BYOK) supported; AI calls use a zero-data-retention API.

Pros

  • Built around email and Word as the core way of working, fitting into lawyers' existing processes
  • Agent output is returned as reviewable .docx and .xlsx files, which lawyers review before approving
  • Every redline is tied to a firm-defined playbook rule, with firm-wide usage analytics
  • The official website states it is SOC 2 Type II certified and compliant with ISO 27001:2022, GDPR, and HIPAA
  • AI calls use a zero-data-retention API, so prompts and outputs are not stored or used for training by the model provider
  • Offers matter-level data isolation, exportable audit records, a zero-retention mode, and US/EU data residency options

FAQ

The official website states that customer data is stored in mixus infrastructure that is SOC 2 Type II certified, with HIPAA attestation; AI calls use Anthropic's zero-data-retention API, so prompts and outputs are not stored by the model provider and are never used for training.
The official website states that mixus can also generate tracked changes; the difference lies in enforcement: mixus ties every redline to a specific playbook rule defined by the firm, learns over time from reviewers' accept/reject decisions, and gives partners firm-wide analytics on playbook usage.

Tool Information

Developermixus
Release Date8/31/2026
Pricingpaid