When It Finally Clicked
Recently I sat down with Helen Fan for an episode of the podcast. She’s the founder of Open Claw LLC, a AI native law firm experiment that went viral, and the author of the Legal AI Value Stack, the piece that travelled through the legal AI world earlier this year and made a lot of people rethink where defensibility actually lives in this market. Something clicked for me in that conversation, and afterwards I sat down and wrote down what I have been boiling over for the past few months.
Most of the AI-native conversation today is about firms that started from scratch, such as Lawhive, Eudia, the YC W26 batch, or about solos who can pivot a practice overnight. That’s a realistic path today towards AI native, but it isn’t the path most 5 to 30 lawyer boutiques are on. The boutique with real clients on the books, two partners who’ve heard the AI-native conversation a few times now, and a team that’s been using ChatGPT seats for six months without much direction, that’s the reality for a lot of the boutiques today and the climb towards AI native looks pretty different.
Data Doesn’t Come First
The standard advice for AI, and the one I have been guilty of giving, has been to get the data in order first, then put AI on top of it - garbage in, garbage out, the more high quality data you feed any system the better what comes out. Once your information is clean and structured and discoverable, the AI can do something useful with it.
While the logic here is rock-solid, the reality is that for most boutiques, getting that data in order is an enormous project. Years of information scattered across the DMS, the CLM, inboxes, SharePoint, personal laptops, and sticky notes on people’s desks, plus all the process knowledge sitting only in people’s heads. The simple test for whether a firm is “data ready” is this: if you ask an associate to pull comprehensive information about a client, do they come back in minutes or hours? Exactly.
Trying to fix all of that before deploying any AI is what makes you quietly give up in month four.
The order of operations in Helen’s Legal AI Value Stack inverts the conventional view. The data infrastructure doesn’t come first; it emerges from the workflows themselves, accumulating one workflow at a time. The structured client info you capture in intake becomes what matter management runs on next, the matter structure that gets built in matter management becomes what billing pulls from, and each workflow you finish leaves the next one with less work to do because the data context it needs is already in place.
For a boutique without dedicated AI headcount, this inversion is the thing that makes the climb realistic. Three workflows shipped one at a time over a handful of weeks each, a data layer building underneath as a consequence of doing the work, and deeper AI applications opening up after that. None of it requires venture money or rebuilding the firm from scratch.
Where The Climb Starts
The way to approach this is to divide the firm into three functions first - intake, matter management, and billing. These three are where most of the non-billable capacity at boutiques is being absorbed today, and they’re where you don’t need AI doing anything close to legal judgement. Get them running, and you’ve bought yourself room to look at the more complex parts of legal work where AI starts getting interesting.
There’s also a purpose to the order of things here. Intake captures the client structure first, which is what matter management needs to do its work. Matter management then adds matter structure on top - status, deadlines, work product references, all of which only makes sense once the client structure is in place. Billing comes last because by then both the client and matter layers exist, and the financial structure can stitch into them cleanly.
The article walks through each of the three workflows in detail, including what the operational move tends to look like for each one and where the diagnosis points elsewhere. For most boutiques in this segment, the moves tend to be lighter than people expect - most firms already have the systems and already have partial records inside them, and the work is closer to plumbing than rebuilding.
The Boutique Law Firm Advantage
In eighteen months, the operational difference between a firm that’s gone through this climb and one that hasn’t is going to be visible - in capacity, in client conversations, in how partners spend their time. That’s the real stake, and I’ve seen enough firms now to have an honest opinion on whether it’s true.
This kind of work is also counterintuitively more achievable inside a boutique than inside bigger firms. Fewer lawyers, fewer decision layers, so that one person, whether it’s a partner, an ops lead, or an advisor brought in for a defined window, can genuinely hold the whole picture and push it through. A 2,000-lawyer firm cannot produce that person by design. A boutique can. That’s one more argument for why boutiques are actually better positioned to punch above their weight on AI than most of the public discourse gives them credit for.
Where The Climb Lands
The climb that fits a 5 to 30 lawyer boutique looks like this: three workflows shipped one at a time, a data layer building underneath as a consequence of doing the work, and deeper AI applications opening up after that. None of it requires venture money or rebuilding the firm from scratch.
The full article goes deeper, including a walkthrough of intake, matter management, and billing with the same template each time, the stitch/buy/build choice underneath each one, and the questions I haven’t figured out yet - what happens when one partner is in and the other isn’t, where the framework probably stops fitting past thirty lawyers, and where stitching breaks down over time.
Legal AI in Action
🎬 Before You Buy Another AI Tool, Try This Sequence
What two steps to take before even picking between the four moves, the three types of walls firms hit when they push their existing setup, and when each of buy, stitch, build, or hire is genuinely the right call for a boutique firm.
🎙 What 1,300+ Hallucinations Tell Us About Legal AI
Damien Charlotin is an arbitration lawyer, data scientist, and lecturer whose AI hallucinations database has become the most referenced resource on the topic in the legal world. We talked through what 1,300+ hallucination cases reveal about who is actually getting caught (not who you’d think), why specialized legal AI tools can sometimes produce more hallucinations than the generic ones, how AI is opening up kinds of legal work that used to be too expensive or time-consuming to attempt, and why the predicted consolidation of the legal market probably isn’t what’s coming.
Coming up!
🎙 Next Tuesday at 2pm CET!
Next week’s guest on Rok’s Legal AI Conversations is Jiyun Hyo, the founder of Givance, building an MSO structure that runs the operational layer of law firms, instead of selling them AI tools.
We discuss why he made that pivot, what AI native has to actually mean to be more than marketing language, and why the operational layer, admins and paralegals, is where AI adoption actually pays off inside a firm.
Each edition of Legal AI Brief brings practical lessons from firms using AI safely.