Architects Journal - New practice Arka Works: ‘Ignoring generative design and AI is not an option’

September 7, 2023

No items found.

This article was originally published in the Architects Journal.

Arka Works operates with one foot in traditional practice and one in software development

The latest in a series of practice profiles looking at architects who have recently decided to go it alone

Name Arka Works
Founded Summer 2023
Main people Keir Regan-Alexander

Tell us about the practice and where you’re based

Arka is a creative consultancy operating with one foot in traditional architectural practice and one in active software development.

It’s just me for now - working as a consultant to architecture practices, clients and software start-ups building new tools for designers, so I’ll be wherever they are.

Most of my network is London-based, but the start-ups are much more wide-reaching – in the US and, increasingly, Oslo, which has become a bit of a melting pot for generative design.

We expect automation to become ubiquitous

Arka Work’s mission is to champion good design judgement and high-quality placemaking in an era where we expect automation to become ubiquitous.

We’ll do this by focusing energy on the practical adoption of new tools that we feel are being built to solve the right problems. We want to concentrate on the technology that will really augment the way we work as a profession and empower architects, as a counterpoint to tools that are clearly looking to replace the more fundamental design work of architects and designers.

We don’t think ignoring emergent generative design & AI is an option for architects, we need to be very engaged.

We’ll also be helping practices manage changes in the way they work over time with training and knowledge sharing. All of the work will come from the practical perspective of having weathered many storms as a leader in an AJ100 practice.

Where have you come from?
I’m founding Arka after 10 years at Morris+Company, where I was a director designing and delivering projects in housing, workplace, health and education.  Before that, I was at Feilden Fowles and AHMM.

Outside of practice I’ve had a growing interest over the years in AI and tech and have been looking for ways to combine them with architecture. I took evening classes in Python and data science during Covid and built some very basic programmes with more talented friends.

Earlier this year I teamed up with two others on a piece of software that produced operational energy assessments specifically for the electrification and retrofit of domestic homes. This allowed me to apply my Passivhaus knowledge of building fabric to a simple and accessible tech product idea.

We didn’t go for funding in the end, but we made something I felt was great, and I now have access to software that can produce Passivhaus planning packages (PHPP) very quickly from 3D models and test lots of retrofit options.

This process demonstrated to me what was possible when architects work directly with software developers – and it gave me the idea for Arka.

Early design concept for a courtyard house project, image produced in AI art generator Midjourney

What work do you have and what kind of projects are you looking for?
We’ll be working with small and medium-sized practices that don’t have much in-house tech R&D. We bring large client and project delivery experience in house and we’re now partnering as a UX Advisor for Preoptima, which is a UK start-up focused on using AI for live carbon prediction as a design driver. At the same time, we’re taking meetings with 'generative-curious' clients and developers that we hope to collaborate directly with soon.

We’re about to start doing a series of training workshops with practices in London to demonstrate what can be done with the new toolset. When I meet clients, I tend to find they are all looking to apply the potential of AI in slightly different ways depending on their sector specialism.

I’ve heard some novel ideas, such as commercial retrofit for upcoming EPC upgrades, rapid carbon prediction to assess demolition alternatives, live testing of different housing prototypes at a masterplan scale, and building compliance co-pilots to help with the Building Safety Act. These are all ideas that could be built and that we wouldn’t have been considering a year ago.

What are your ambitions?
Things are moving fast and six months feels like a long period of change at the moment. We will go wherever we can be most impactful. I’m planning to run this like a start-up, to stay lean and pivot to where our clients tell us there is value.

What are the biggest challenges facing you as a start-up and the profession generally?
When I speak to practice leaders there’s a growing consensus that it’s increasingly hard to make traditional models of practice stack up financially.

The profession is more competitive than it has ever been, and I’ve seen overhead costs just going up every year. At the same time, the regulatory environment is burdening more responsibility on to architects without financial reward. After Brexit and Covid, the projects are fewer in number and there is also a really toxic trend for architects devaluing themselves, which is reflected in the fee environment.

There’s a toxic trend for architects devaluing themselves

While conditions in practice are tougher than I’ve seen since 2009, I see this new wave of tech bringing the potential to release creative minds and shake up the way we work for the better.

I don’t think there is a paragon of practice model that we should be looking to preserve, because the prevailing mode doesn’t work brilliantly for the majority of architects anymore.

It feels as though the time is right for a shift in gear, and automation could actually be the catalyst for a more creative style of practice, allowing architects to focus on doing the ideas-based work that brings the most value and letting software take care of the repetitive taskwork that no one enjoys.

Is there more scope to experiment with new technology as a start-up?
Yes, I hope we can be incredibly agile. We don’t have to talk about things, we can just get on and do them.

Which scheme, completed in the last five years, has inspired you most?
The last time I felt really moved by architecture was watching a film during lockdown about Charles Bello. He is a Californian architect who has lived off-grid in a huge remote forest which he planted 55 years ago.  He talks about trees like people and he is an incredible carpenter and tinkerer; hacking together traditional and new methods in very experimental ways and making everything from scratch from his sawmill. This is a true vision of innovation in my mind.

Over the decades, he has built several very humble, but gorgeously crafted homes for his family in the grounds, each one a bit more ambitious than the last.

The last structure he built really sticks in my mind: a handmade circular gallery with a filigree timber structure to house his many wooden sculptures. It is an absurdly accomplished and well-poised piece of architecture.

How are you marketing yourselves?
By learning in public and being honest about what works and what doesn’t. I’m very open about the research and experimentation that I’m doing and I try to write it up whenever I can.

I find that writing is like sending a boomerang into the unknown. You don’t know exactly how it will return but it always does.


Midjourney AI tool experiment on natural tidal pools by Arka. The concept is inspired by the forms of Eduardo Chillida, blending organic and human-made form

More like this

I have begun to think that Generative Design was waiting for Generative AI to come along to really find a place in the mainstream.

GenDesign Vs GenAI: Know Your Apples

I have begun to think that Generative Design was waiting for Generative AI to come along to really find a place in the mainstream.

October 30, 2023