What it is
Streamline AI is the front door in-house legal teams install when they want prebuilt legal request types rather than a canvas to configure. Requests arrive through Slack, email, Salesforce, or a web portal; conditional logic on the form surfaces only the fields that matter to that request type; a routing rule assigns it; and a matter record opens behind it carrying the documents, tasks, approvals, and reporting. Founded in 2020 by Kathy Zhu (CEO, previously Assistant General Counsel at DoorDash) and Julian Wimbush (previously a product lead at Google), the company raised an $8.6M Series A led by Blumberg Capital on 30 July 2025 — with Tribeca Venture Partners, Acronym Venture Capital, Great Oaks Venture Capital, and Scribble Ventures participating — taking total funding to $14M. Named customers include Gusto, 8x8, Bloom Energy, Horizon3, VSCO, Branch Metrics, and Pantheon.
Why it shows up in Legal Ops stacks
- It ships legal request types, not a builder to make them. Prebuilt intake covers contract review, NDAs, vendor agreements, marketing compliance, and privacy requests on the day you sign, and legal ops adjusts rules and routing without involving IT. The vendor’s own comparison page states the product typically deploys within 60 days, with many teams live in weeks. That is the entire trade: narrower range than a general workflow engine, in exchange for not designing your own intake taxonomy from scratch.
- The April 2026 release put two named agents on top of the workflow layer. The In-House Legal AI Platform launched on 14 April 2026 with Velo Copilot, a coordinating agent that works across matter types to apply institutional knowledge, surface risk, and recommend or start the next step, and Featherline, a contract review agent that applies negotiation playbooks, flags risk, and reports on contract metrics. Streamline publishes acceleration of more than 40% and roughly 20% of legal capacity freed at customers including Gusto and 8x8. Both figures are vendor-published and unaudited.
- It is callable over MCP, which is rare in this category. An MCP Connector reached general availability on 15 July 2026 alongside a Slack Intake Agent that opens a structured request from a natural-language description. The connector authenticates with OAuth 2.1, enforces each user’s existing Streamline permissions, answers without storing or caching customer data, and writes every interaction to the audit trail — so an employee can ask Claude, ChatGPT, Gemini, Microsoft Copilot, or Glean for a matter status instead of interrupting a lawyer. See legal intake and matter management for the underlying model.
Pricing reality
Streamline publishes starting prices, which most of this category does not. Pro starts at $22,900 and Enterprise at $26,900, each including four core users and adding SSO, SCIM, Slack, Microsoft Teams, and e-signature through DocuSign or Adobe Sign. The $4,000 step to Enterprise buys the integrations that decide whether this becomes your system of record: storage (Google Drive, Box, OneDrive, SharePoint), a Jira custom channel, Salesforce, CLM through Ironclad, NetDocuments, and API access. Pro caps business users at 5,000; Enterprise removes the cap. Knowledgebot, Dynamic Docs, and regional data hosting are priced separately.
Read those numbers as a platform floor, not a budget. The page prints no billing period and every route ends at Contact Sales, so the published figures set the entry point and the add-ons set the actual invoice. Four core users is a small legal team — a five-lawyer department is already buying seats.
Best for
Legal ops managers and GCs at companies of roughly 200 to 3,000 staff where legal is drowning in high-volume, repetitive commercial requests and nobody has three months to design a workflow taxonomy. The trigger: contract review and vendor requests arrive in Slack and email with no queue, the team cannot report demand by request type, and the business is asking for an SLA legal cannot yet measure. It is also the sharper pick when you want legal data reachable from the AI assistant the company already runs.
Versus the alternatives
Checkbox is the direct rival and the honest comparison: Checkbox is a no-code engine that will also run compliance, HR, finance, and procurement intake in the same tenant, and Streamline is legal-only and faster to stand up. Pick Checkbox when four departments want one front door; pick Streamline when only legal does and time-to-live is the constraint. Onit is the enterprise legal management incumbent — take it when matter management and outside-counsel spend control are the real purchase and intake is just the form feeding them. Ironclad is the pick when the contract is the object of record and request capture is a CLM stage rather than the product. Wordsmith is the fastest-growing entrant in the segment and the better fit when you want legal reasoning callable over MCP without buying an intake system underneath it. GC AI fits when the lawyers want a better assistant and business-side intake is not the problem you have.
Watch-outs
- Legal-only scope is the reason it is fast and the reason it caps out. When procurement or HR later ask for the same front door, the answer is a second tool rather than a second workflow. Guard: ask during evaluation which non-legal request types the platform will accept, in writing, and price the two-department scenario before you sign the one-department contract.
- Velo Copilot and Featherline are new enough to have no independent track record. They shipped in April 2026, and the efficiency numbers attached to them are the vendor’s own. Guard: instrument the baseline before go-live — requests per quarter by type and channel, time to first response, share of contract reviews closed without a lawyer editing — and gate renewal on a number you measured, not the 40%.
- The published price is a floor with four seats and metered add-ons above it. Knowledgebot, Dynamic Docs, and regional data hosting all carry separate pricing that the site does not print. Guard: get a quote with year one and year three side by side at double the core users and your real request volume, add-ons included, in one document.
- Regional data hosting is a paid add-on, and the AI data terms are not public. Streamline holds SOC 2 Type II, hosts on AWS in ISO 27001 certified data centres, and encrypts with AES-256 at rest and TLS in transit, but the public site names no model provider, no default hosting region, and no commitment on whether customer content reaches model training. Guard: get the subprocessor list, the model provider, and a written no-training clause into the DPA before privileged material passes through, and price the hosting add-on explicitly if you are an EU entity under GDPR.
- An MCP connector widens who can reach legal data. Permission-aware and audit-logged is the right design, but the population asking questions goes from the legal team to everyone with an AI assistant. Guard: confirm the permission model against your own roles during evaluation, and review the audit trail after the first month rather than assuming defaults matched your matter confidentiality rules.