We help product teams embed AI into their products

At Superflows, our mission is to give everyone access to an expert in the software they're using.


In the business world, powerful software tools are everywhere. They’re typically built with large feature sets and the flexibility to adapt to a wide range of use-cases. These two properties add complexity to the software and require a large numbers of screens. In turn, this makes the software more difficult to learn and use.

Starting off with this software is often overwhelming. You’re funnelled down a tacky onboarding flow which highlights buttons on the screen. These are instantly forgotten. You came to the software to solve your problem, but now you’re being lectured at instead.

For the product managers involved in designing the software, this looks like poor activation and retention. They know the features that deliver value, but not enough people reach them. They build out extensive documentation, run training seminars, and revamp the onboarding flow, but the metrics don’t budge.

We’ve felt this problem from both sides. In the early days of our company, we pivoted a lot. We had to learn to use a lot of new software as our needs changed, and the products we built often had poor activation and retention.


The solution to this problem is to add an interface to the software that is so easy to use that anyone can use it to solve their problem immediately. Even if you don’t know how to solve your problem in the new software, you can at least explain what you’re trying to do. Now imagine if you had an expert on hand to ask questions to.

That’s why we’re building Superflows: to enable any software product to embed an AI product assistant into their software.

Large Language Models (LLMs) make this possible. So we’re building the open source tooling to make using them to build a product assistant easy and available to everyone.


We’re a small team, but full of energy. Software Engineer James Rowland started out coding by controlling high-powered lasers for neuroscience experiments in his PhD at Oxford. CTO Henry Pulver comes from a machine learning and software engineering background by study at Cambridge and background at FiveAI. And CEO Matthew Phillips completed his neuroscience PhD from UCL 2 years ago.

We work in person every day in London. We all share a love of building products that solve problems for people and make the world work better. Get in touch with us if you share our mission or would like to use Superflows.

  • Henry Pulver

    Co-founder & CTO

    AlphaGo inspired me to work on machines that can act intelligently - so I wrote my Master's thesis at Cambridge on reinforcement learning for Atari games. Following that, I worked at FiveAI on autonomous driving research as a machine learning research engineer, before leaving to start Superflows. Outside of work, I enjoy watching Arsenal FC and playing sports.

  • Matthew Phillips

    Co-founder & CEO

    I started out thinking I would become a scientist. After publishing some papers I found that my academic itch had been scratched, and I wanted to get closer to where the value is delivered. To me, that means increasing the ability of people to access technological advances. I live in London and my spare time is usually spent reading - often about science and tech - or playing with my two dogs, Kiwi and Goose.