The software that decided the legal department was never the bottleneck. The inbox was.
It is 9:14 a.m. at a mid-size tech company. A sales rep needs an NDA reviewed before a noon call. A year ago that request would have landed in a general counsel's inbox, fought for attention against forty others, and surfaced sometime after lunch - maybe. Today it enters a form, gets routed by rules the legal team built themselves, lands with the right reviewer, and shows up on a dashboard the moment it is filed. Nobody chased it. Nobody forgot it.
That quiet, unremarkable piece of order is what Streamline AI sells. The company builds an AI-powered intake, triage and matter-management platform for in-house legal teams - the lawyers who work inside companies rather than at law firms. More than 500 of them now run their daily work on it. It is not glamorous. It is the opposite of glamorous. That is rather the point.
Here is the uncomfortable truth about in-house legal work. The lawyers are good. The processes are medieval. Requests arrive by email, Slack, hallway ambush, and the occasional forwarded thread with twelve people on it. Work gets tracked in spreadsheets, or in someone's memory, which is worse. When the CFO asks how the legal team is doing, the honest answer is usually a shrug.
Kathy Zhu lived this. As the first commercial counsel at Medallia and then DoorDash, she was the person every team funneled requests to as the company scaled underneath her. The volume was relentless. The tools were not built for it. And the legal department, doing genuinely hard work, kept getting cast as the slow lane - a reputation built less on reality than on the simple fact that nobody could see what legal was actually carrying.
In 2020, Zhu teamed up with Julian Wimbush, a former Google product lead and engineer, to build the tool she had wanted and never had. Their bet was contrarian: that the problem with legal departments was not the lawyers and not even the workload, but the absence of a workflow layer the rest of the company had taken for granted for a decade. Sales had Salesforce. Engineering had Jira. Legal had an inbox and good intentions.
The wager came with a reframe. Zhu's pitch was never "make lawyers work harder." It was "stop making legal invisible." Give the team a way to capture every request, route it automatically, and - the part that turns heads in the C-suite - produce real numbers about turnaround, volume and bottlenecks. A legal department that can show its work stops being the department of no and starts being a measurable partner.
The platform's first trick is the least flashy and most important: a no-code builder that lets a legal team design its own intake forms and approval workflows without filing a ticket with IT. Requests come in structured. They route themselves. Nothing falls into the void of an unread inbox. Around that core sits the rest - AI that reads inbound email and extracts the request automatically, matter management that tracks each item across its full lifecycle, dynamic document generation, a conversational knowledge bot, and analytics dashboards that finally answer "how is legal doing?" with a chart instead of a shrug.
In April 2026 the company launched what it called a first-of-its-kind AI platform for in-house legal work, leaning harder into automated email intake and AI-assisted review. It plugs into the tools teams already live in - Slack, Salesforce, email, and contract systems like Ironclad - because a workflow tool that demands you abandon your other tools is just another inbox.
No-code custom forms with automatic routing and prioritization, so requests stop getting lost.
AI reads inbound email, pulls out the request, and creates the matter for you.
Track every request across its full lifecycle, with visibility, audit trails and access controls.
Real-time metrics on turnaround, volume and bottlenecks - the legal department's story, in numbers.
Four features that sound boring at a dinner party and save an hour per request before lunch.
Kathy Zhu and Julian Wimbush start Streamline AI to fix the intake chaos Zhu lived as first counsel at Medallia and DoorDash.
Legal departments at growth-stage companies adopt the no-code intake and workflow tooling; the customer roster grows.
Round led by Blumberg Capital, with Tribeca Venture Partners, Acronym, Great Oaks and Scribble Ventures. The platform now supports 500+ lawyers.
CEO Kathy Zhu makes the public case that legal departments win by measuring and reporting their own impact.
Streamline AI launches a refreshed, AI-forward platform billed as first-of-its-kind for in-house legal work.
Self-reported by customers and the company. Treat as directional, not audited - but directionally, that is a lot of recovered Tuesdays.
Proof, in this market, is a customer list other lawyers recognize. Streamline AI's includes the legal teams at Gusto, 8x8, Acorns and Bloom Energy, with health-and-consumer names like Cityblock Health and Hims also in the fold. The 8x8 deputy general counsel has put a number on it: 30 minutes to an hour saved per request. Multiply that across a queue and you have bought a legal team back a meaningful slice of its week.
The money agreed. Blumberg Capital led the $8.6M Series A in July 2025, bringing the total to roughly $14M, with Tribeca Venture Partners, Acronym Venture Capital, Great Oaks and Scribble Ventures along for the round. Among the backers are legal heavyweights, including former Facebook general counsel Colin Stretch - people who have personally felt the inbox problem from the inside.
Strip away the product specs and the mission is almost stubbornly simple: equip in-house legal teams with the tools and the metrics to drive outsized impact. The vision underneath it is the part that gives the company its edge - a future where legal teams are seen as accelerators of business results rather than the place deals go to wait. It is a reputation rehab project disguised as software.
There is a pleasant irony here. The company building AI for lawyers is led by one - Zhu is co-founder, CEO and general counsel, which means she is also, technically, her own demanding customer. A product built by the person who has to live with it tends to skip the features nobody asked for.
The legal AI gold rush is loud right now, and most of the noise is about drafting and research. Streamline AI is making a quieter bet on the unsexy layer beneath all of it: as AI generates more contracts, more requests and more documents, somebody still has to route, prioritize, track and measure the flood. The teams that have a workflow system will absorb the surge. The teams still living in their inbox will drown in it. The tooling that looks like plumbing today becomes load-bearing tomorrow.
So back to that NDA at 9:14 a.m. The version of this story from a few years ago ends with a frustrated sales rep and a lawyer apologizing for a backlog nobody could see. The version running on Streamline AI ends before lunch, on a dashboard, with a number attached. Same lawyers. Same workload. Different system. The legal department did not get faster by working harder - it got faster by finally being visible. That is the whole bet, and so far it is paying out.
Sources: streamline.ai, LawNext, SiliconANGLE, Finsmes, BusinessWire, Legal Dive, Crunchbase, Law.com.