AI consultancy
Finding where AI will make a real difference to your business. We spend time with the people doing the work, spot the best opportunities and tell you plainly if AI isn't the answer.
Learn more
AI & Emerging Technology
We stay close to emerging technology and turn what we learn into working software. Much of that is now AI — production features, agents and internal tooling, website AI readiness — and some of it is not, from real-time 3D and augmented reality through to Gaussian splatting. The common thread is quickly getting past the demo to something in production that adds value to your business.
From Opportunity to Prototype
We start by finding the work where AI would change the economics: the questions your team answers the same way forty times a week, the report someone rebuilds by hand every Monday morning.
That starts with us in the room. We spend time with the people actually
doing the job, because the opportunities worth chasing are specific to your business and nobody knows your business like your people.
Then we build the most valuable ones quickly, inside a few weeks, running
against your data rather than a curated sample. We build prototypes on the same stack the real thing would use, so it is a foundation rather than a throwaway. We iterate with you to quickly reach real solutions for your business.
AI Readiness
The question we are asked most at the moment is what AI means for a website. People increasingly arrive at your content through an assistant rather than a search engine. There is a set of things to get right here and we know what they are. Some of it is classic good engineering, which we have been doing for years. Some is new, and we have done the work to find out what works and what doesn't.
Beyond the website, the answer is not always to deploy AI directly. Just as often it is that AI now lets us build traditional tooling, quickly: something hyper-specific to one team's way of working, the kind of thing that was never worth a three-month budget and so never got built. That is the part we think is most underrated at the moment, and a great deal of useful software is now economically viable.
AI Isn’t the Only Game in Town
Not all of the interesting technology is AI. We have been building with tools on the bleeding edge of technology for over a decade, and we are always looking out for new ways to leverage interesting developments to help our clients.
Gaussian splatting, augmented reality, real-time 3D, on-device models, voice. Being early to these is easy and not worth much on its own. The value is in knowing what each one is actually good at, what it costs to run and what it is still bad at. When we suggest something for your project, it is because it solves a problem you have, not because it is having a moment.
Finding where AI will make a real difference to your business. We spend time with the people doing the work, spot the best opportunities and tell you plainly if AI isn't the answer.
Learn more
AI & Emerging Technology projects
- Fintech & BankingUX & Product DesignMachine Learning & AI
QuickBooks
Since 2022, we've helped QuickBooks understand its customers and turn more website visitors into sign-ups. One project became many, and we're now a go-to partner for their international marketing team.
- Food & DrinkUX & Product DesignBrand & Visual Design
Deliveroo
We consistently deliver for our long-term partner Deliveroo, helping various teams across their business with both marketing and internal tooling.
- SportsMedia & EntertainmentAudience Research
New York Yankees
The New York Yankees wanted to make the fan tour of Yankee Stadium more memorable. We built a game that brings fans closer to their Yankees heroes.
Common questions on "AI & Emerging Technology"
Where should we start with AI?
With the work, not the technology. We spend time with the people actually doing the job and look for the things that happen over and over: the question your team answers the same way forty times a week, the report someone rebuilds by hand every Monday morning.
Those opportunities are specific to your business and they are rarely visible from an org chart. Once we can see them, it is usually obvious which one or two are worth building first.
How long until we see something working?
Weeks, not quarters. We build the most valuable opportunity quickly and run it against your data rather than a curated sample, because a demo on tidy example data tells you almost nothing about whether it will hold up.
From there we iterate with you. Seeing the thing half-working in your own context is what makes the next decision easy.
Is a prototype something we then have to rebuild?
No. We build prototypes on the same stack the real thing would use, so what you end up with is a foundation rather than a throwaway.
If the prototype proves the idea does not work, you have found that out in a few weeks for a few weeks of cost. If it works, you are already partway to production.
Do we need our data in order before we start?
Usually not. Waiting for a data platform to be finished is one of the most common reasons nothing ever gets built, and most useful projects need one messy source to be good enough rather than everything to be perfect.
We would rather start with what exists, find out what the work actually needs, and let that tell you which parts of your data are worth investing in.
What does AI readiness mean for our website?
People increasingly arrive at your content through an assistant rather than a search engine, and that changes what a site has to do: your content needs to be legible to a machine that is answering on your behalf, and accurate enough that the answer it gives is one you would stand behind.
Some of getting that right is classic good engineering, which we have been doing for years. Some of it is genuinely new, and we have done the work to find out what helps and what does not.
What if AI is not the right answer for us?
Then we will say so. Just as often the answer is that AI now lets us build the tool itself quickly: something hyper-specific to one team’s way of working, the kind of thing that was never worth a three-month budget and so never got built.
That shift is the part we think is most underrated. The threshold for what is worth making has dropped, and a lot of useful software sits just under where it used to be.
You mention things beyond AI. When is that relevant?
We have been building with bleeding-edge tools for over a decade: Gaussian splatting, augmented reality, real-time 3D, on-device models, voice. Being early to these is easy and not worth much on its own.
The value is in knowing what each one is genuinely good at, what it costs to run and what it is still bad at. When we suggest one for your project, it is because it solves a problem you have, not because it is having a moment.
Want to chat about an AI project?
Give our founder, Iggy a shout...