Don't Call It a Comeback

Tasuki (襷) is a Japanese word, and the cord it names has two uses. In the ekiden, Japan's famous long-distance relays, the runners don't pass a baton. They pass a tasuki, a cloth sash worn across the shoulders, handed from one leg of the race to the next, and the phrase for taking your leg is to inherit the tasuki: the team's effort and trust, carried forward. The rest of the time the same cord ties back your kimono sleeves, freeing your hands for the hard work.

You saw the name earlier this season on a bus shelter in San Francisco, a URL on a white wall with a little game waiting behind it. It fit a stunt like that, partly because it sounds like one, somewhere near Sudoku, and partly because I already owned the domain. But I bought that domain for what the word means, not how it sounds.

This is what it's for, both halves of it. Most of what's broken in a business lives in the handoffs, the places work passes from one person to the next, and the first thing you do in a process workshop is trace exactly that: who has it, who gets it next, how it gets there. That's the relay. And the AI tooling is the cord that ties the sleeves back, clearing the busywork off your hands so you can do the part that actually matters.

Where I got ahead of myself

What I got wrong at first was who delivers it. I'd been living on the bleeding edge of agentic everything, so my answer was an agent: Tasuki would join your Slack, run the whole thing from inside the channel, no human in the loop. A bot with a job.

Then I spent months actually deploying this stuff, real products, real customers, and the lesson landed. It's going to be a while before a company hands an agent the keys to anything that matters. Agents don't have the hall pass to the enterprise yet. Consultants do, the people who sit in the room, carry the trust the tools haven't earned, and change the work alongside you. The line I landed on: the human interface to AI in your business.

I've seen this shape before. When SaaS arrived, the companies that got there early didn't buy software off a shelf, they brought in people who knew how to make the new thing work. I was one of those people, an early Salesforce partner back when half the market still didn't trust the cloud. Same job, new wave. So don't call it a comeback. I've been here for years.

Rolling up my sleeves

What moved me from believing it to building it was watching Anthropic put real weight behind professional services, a public bet that this technology reaches businesses through people, not just a product. That's the bet Tasuki is built on, so I paid attention.

So I went through their learning path, the courses they put in front of anyone who wants to build a practice on their stack. I did it for two reasons: to pressure-test what I'd been building on Dropsignal and Bordmappe against what the people who make the tools say you actually need to know, and because those two products are how I'd show the work. Four courses, four certificates.

I put them on my LinkedIn, which I have never once done in my life. They're step one of that path, the part they ask for before anything else, so posting them was less a credential than a signal. Step one, done.

The studio

Everything you've watched me build this season is the pitch. The car pages and the city screens were never side quests. They are the portfolio, and the proving ground too, the place I figure out the stack and the deployment before any of it goes near someone else's business, and the place I'd ramp anyone who came to build alongside me. Three names, one studio, and the seams between them are not luck, they are the work. Turning a one-off proposal into a product with a name is a deliberate move, the one I keep making on purpose.

I'm not telling businesses I could bring AI into their operation. I'm showing them I already do, on my own dime and in public, an episode at a time. Tasuki is the door with a handle on it. The other two are what's on the other side.

And the capability underneath isn't a toy. What Bordmappe does, reading a messy stack of service records and pulling out the few things that actually matter, is a smaller version of a problem I sat with as a CTO years ago: twenty thousand leases that needed abstracting, and a consulting firm's quote to do it that ran to a number I still don't believe. The leases were gnarlier than any car's file, but it's the same shape of work, and the army it used to take is now a tool.

Getting the band back together?

For now the only face on this is mine, and that shouldn't last. The more I sit with Tasuki, the more I think about the people I've solved problems with across a whole career, the ones beside me on the early enterprise rollouts and the practices we built, the ones we won and the ones we badly lost. People who've sat in every seat there is, and watched more than one wave come in.

The job itself never changes: solve the problem with the tools and the skills you've got. What's new is that nobody has written the playbook for doing it in this age, because the age is too young to have one. The people best placed to write it are the ones who've stood up the strange new thing before.

So the invitation is loose on purpose. Go through the learning path with me, take the baseline of the stack I build everything on, and let's write the playbook nobody's handed us. Less a job posting than a standing invitation to jam.

If you read all of this and felt it, if you've lived some version of it, come find me at christian@tasuki.co. I'd love to hear it.