I was invited to a law firm in Nairobi to talk about how we could work together. The managing partner and I had agreed on a time. I showed up on time, which is almost always the wrong strategy in this country.
The secretary apologised. An urgent matter had pulled the managing partner away. An hour, she said. Maybe a little more. Would I mind waiting? She offered me a seat and asked if I'd like to walk around the firm in the meantime.
I'll save you from the actual number of hours I waited. This is Kenya. You know how it goes.
Here's something about me, though. I sometimes wonder if I picked the wrong career. I think I'd have done well in the DCI. Maybe even as a lawyer. Because the thing I enjoy most, more than almost anything else in my work, is sitting quietly in a room and listening. Not the polished version of what people tell you — the version that comes out when they forget you are still there.
So I sat. And I watched. And I listened. And by the time the managing partner finally walked in, I already knew more about her firm than she was about to tell me in the meeting.
What I went there to pitch
I had come to present SISU by Lenhac — our law firm management system. Case management, billing, client portal, the whole thing. I had the demo ready. I had the pricing ready. I had the opening line warmed up.
Within the first thirty minutes of walking around the firm, I quietly put the pitch away.
The registry was a mess. Not a "needs tidying" mess. The kind of mess where closed files had genuinely ended up on the floor. Where a few things had been scanned at some point, by somebody, and nobody was sure where the scans now lived. Where advocates — qualified, called-to-the-bar, billing-at-lawyer-rates advocates — were walking upstairs and downstairs looking for files they would probably not find.
And I kept asking myself the same question as I walked around. Why is a trained advocate spending two hours of her day looking for a document she does not even need? She needs the content. Not the paper. The content.
Nobody in that firm was lazy. Nobody was careless. It is just how law firms end up after years of turnover, overstretched associates, and matters that pile up faster than anyone has time to file them. If you own a law firm in this country, or you work in one, you already know what I am describing. You have probably stopped noticing it because you see it every day.
What the registry actually looked like
If you have ever worked in a Kenyan law firm, or you own one, you already know what I am about to describe. And if you are a managing partner reading this, I want you to be honest with yourself for the next two minutes, because nothing I'm going to say is new to you.
After a case closes, the files don't really go anywhere. They get moved. Sometimes to a shelf. Sometimes to a room. Sometimes, genuinely, to the floor. I am not exaggerating. I saw files on the floor.
And look, this is not because the people in that firm are careless. It's because of how the industry actually works. Law firms have some of the highest staff turnover of any professional service business in this country. An associate joins, works for two years, leaves for an in-house role at a bank. Another one comes in. The files from the cases the first associate worked on are still somewhere, but the person who knew where is gone. Multiply that by ten years and you have what I was standing in.
The second thing is that lawyers are overstretched. By the nature of the work. Unless you are the managing partner, you do not have time to file things properly. You have a hearing tomorrow, a client on the phone, a draft due by five, and a bundle to prepare. "Proper filing" is a phrase that sounds lovely in theory and dies quietly every Tuesday afternoon when real work piles up.
So the registry becomes what I saw. Boxes on top of boxes. Some labels. Many without. A few files that had been scanned at some point — somebody clearly tried — but nobody could tell me where the scans were now. On somebody's old laptop? On a hard drive that left with an associate? In a shared drive that nobody maintains? Pick one, they're all bad answers.
The question that kept coming back to me
Here is what I kept thinking as I walked around.
Why is a qualified, trained, called-to-the-bar advocate spending two hours of their day walking up and down a staircase looking for a file?
This is a person who went to law school. Did pupillage. Passed their bar exams. Their time is billable at a rate that would make your eyes water. And right now, at 3 p.m. on a Wednesday, they are on the third floor lifting box lids because a client called asking about a matter from 2019 and nobody can find the file.
Worse — and this is the part that really got me — most of the time, they don't even need the actual physical paper. They need the content. They need to know what was in the agreement. What the other side argued. What the court ordered. The paper itself is almost never the point. The point is the information the paper contains.
And yet there they are, on the third floor.
The cases that stall for no good reason
I talked to one of the associates while I was waiting. She told me — casually, like it was normal, because to her it was — that they had a matter that had been effectively stuck for weeks. Not because of the court. Not because of the other side. Because they couldn't find the file from the earlier phase of the dispute. The context was in the physical file. The physical file was somewhere. Somewhere was not findable.
Think about what that means. A client is paying for a service. The service has stopped. Not because anyone is lazy. Not because anyone is incompetent. Because a cardboard box is in the wrong place in a room.
I have heard versions of this story from every law firm I have talked to in this country, and I am going to say the quiet part loud: this is not a small problem. This is one of the biggest hidden costs in the Kenyan legal industry, and almost nobody is measuring it.
The managing partner walked in
She came in, apologised, sat down, and gave me the opening I expected. The clean, professional, slightly rehearsed explanation of why she was late. The version you give to a vendor.
And here's where I'm going to tell you about one of my favourite tools in a conversation. I don't know if I learned it somewhere or if I just stumbled into it, but I use it all the time. It is the awkward silence.
After she finished her explanation, I didn't say anything. I just looked at her. Not in a hostile way — I wasn't trying to make her uncomfortable. I just… didn't fill the space. And for a moment she looked a bit confused. Maybe wondering if I hadn't heard her. So she started again.
Except this time, it wasn't the script. This time it came out raw. The actual reason she was late. The actual frustration behind her day. The matter she had been stuck on. The file she couldn't find. The paralegal who was meant to have it ready but didn't. The client who was waiting. The court date looming. All of it, coming out the way it actually felt inside her head, not the way it was meant to sound in front of a stranger in a suit.
That is the gem I'm always listening for. Not the well-versed version. The honest one. The details are in the genuineness of the frustration, and you cannot get to them by asking better questions. You get to them by shutting up.
She caught herself after a while, thanked me for being patient, and asked me the question I love most in any of these meetings.
So, what do I do now?
Who wins when I win?
I love that question. Not because it gives me an opening to sell something. I love it because it is the moment I know whether this is going to be a real conversation or a transactional one. And it reminds me of a question I carry around with me all the time, one that I think more people in tech should ask themselves: who wins when I win?
Coincidences are rarely coincidences. Maybe they exist by design. I had just spent several hours wandering around her firm without her supervision, which had given me a view of her operation she didn't even know I had. I could have given her the easy answer. Subscribe to SISU by Lenhac. Case closed. Invoice sent.
I didn't.
Instead, I asked her a different question. If there was one thing you no longer had to do, and your firm could still run efficiently, what would it be?
She didn't even pause. Not a second. She said: I hate the days I have to email the document warehouse on Mombasa Road to retrieve a file. They take forever. Sometimes the file they send is the wrong one.
I had assumed her registry was in-house and chaotic. It turned out it was offsite and chaotic — which is worse, because now you are paying rent on a warehouse as well as losing time. She told me she had tried other law firm systems before and stopped using them, because none of them solved her real problem. Her real problem wasn't case management. It wasn't billing. It was that the context of every matter lived inside a physical file, and the physical file was almost never where she needed it to be.
I told her I did not have a solution for her yet. I needed to sleep on it.
The drive home
I have been deliberately slow on the whole AI-and-law conversation. Partly because I find most of it overhyped. Partly because I have watched too many firms get sold a shiny thing before they were ready for it. You cannot put AI on top of a broken registry. It's like installing a smart irrigation system before you've dug the well.
But on the drive home that day, something shifted for me. I realised that for this firm — and for most firms like it — only AI was going to help. Not because AI is magical. Because the problem was fundamentally about extracting meaning from thousands of documents nobody had time to read, and that is exactly what AI is good at. Everything I'd been resisting was suddenly the only thing that made sense.
We started working together on the problem. That is how SISU AI Case Planner was born.
Watching it work
The first time I watched her team use it, I almost didn't believe what I was seeing. They uploaded documents into the system. The system read them, understood them, and produced a case plan. Not just a summary — an actual plan, with drafts. A plaint. A legal opinion. The beginnings of submissions. Under thirty minutes, start to finish.
She was over the moon. Her team was over the moon. And I was happy, I really was. But for a quiet moment, I was also a little sad.
The part I don't usually talk about
Here is the thing I have been thinking about a lot lately, and I am going to say it out loud because I don't think we say it enough in this industry.
I had been observing her firm for hours before that meeting. And as I watched the AI Case Planner work, I noticed something. The managing partner was barely talking to her secretary anymore. The tasks she used to ask a paralegal to run downstairs for were now happening in a tab on her laptop. The file she used to email Mombasa Road about was now being parsed and summarised in real time. We had just built another tool — one that listens to phone calls and transcribes them into tasks, deadlines, and automatic reminders. More of the small, human jobs that used to sit between her and getting work done were quietly disappearing.
And every time I write code and I can see, clearly, that what I'm building is going to make somebody redundant, I ask myself: is it worth it?
This particular firm, I don't think, is going to fire anyone. That's not her style. But I will tell you what I am almost certain of. When one of her paralegals resigns next year, or the year after, she is not going to replace them. She won't need to. The tool will have absorbed enough of the work that the gap will close on its own. No dramatic announcement. No redundancy letters. Just a quiet decision not to post the job.
Multiply that by every law firm in Nairobi. Then every firm in Kenya. Then every industry.
I don't have a clean answer to the question I keep asking myself. I genuinely don't. I build these things because they solve real problems for real people, and the frustration I saw in that managing partner's face when she finally stopped reciting her script was real, and the relief on her face when the Case Planner worked was real, and I don't think the answer is to stop building. But I also don't think the answer is to pretend there's no cost, or that the cost falls evenly, or that the people who benefit are the same people who bear it.
So what's the point of this story?
Honestly, I'm not sure I'm writing it to sell anything. I think I'm writing it because I want to be honest about what I saw that day, and what I built afterwards, and what I feel about it now that it's out in the world and working.
If you run a law firm in this country and your registry looks like the one I walked through, you already know it does. You don't need me to convince you. You need to decide whether this is the year you finally deal with it. And when you do, please think carefully about the tools you pick, and about the people in your firm whose jobs quietly depend on the current mess. Not because you shouldn't do it. But because someone should be thinking about them, and if it isn't going to be the vendor selling you the software, it will have to be you.
As for me, I'm going to keep building. I'm going to keep asking the awkward question. And I'm going to keep sitting in awkward silences, because that is still where the real answers live.
Comments (0)