In partnership with

Today we meet a startup that has just announced its first funding round.

You might have seen Gendo in the news yesterday when they launched their generative AI product aimed at architects, but today we take a closer look at the product and the company behind it.

– Martin

Intercom for Startups

Join Intercom’s Early Stage Program to receive a 90% discount.

Get a direct line to your customers. Try the only complete AI-first customer service solution.

Apply now

Gendo wants to show what generative AI can do for architects and designers

General purpose AI image generators can be incredibly powerful, but they can also be incredibly frustrating if you’re trying to be productive. 

That specific image you conjure in your head can prove elusive, even after a long session of experimental prompting to try to get anywhere near the perfect result.

For this reason, general purpose AI software can be useless in a lot of professional contexts. And so there’s an opening for specialist tools such as Gendo, which aims to make generative AI useful for the architecture and design market.

“From start to finish in architecture and design projects, you need to convince stakeholders, yourselves, your colleagues, of the design decisions that you're making. And the easiest way to do that is through pictures of what you want to build,” explains George Proud, co-founder and CEO of Gendo.

“The production process of making these images, traditionally, is brute force. It's lots of hardware, lots of technical skills, and lots of time. It can take a week if you're speedy, and cost anywhere from £2,000 to £10,000 to get a render made, whereas we can now do it in a matter of minutes on a browser with no technical skill required.”

By using existing designs created for a project, Bristol-based Gendo can create imagery for any purpose, allowing them to specify lighting, materials, camera angles, and image styles, such as a photo-realistic render or a pencil drawing.

“It's all about specificity and control,” says Proud. “Text to image, text to video, and text to 3D are all useful, but ‘text to anything’ isn't really going to be a design grade solution because you're never going to be able to write the perfect prompt or essay to explain your project.”

How it works

Gendo’s software creates project-specific training data from the customer’s own 3D design work, sketches, photographs of physical models and the like, which they will have created in their initial design and concept work.

Proud says that in addition to this early-stage work, Gendo can also be useful with much later-stage designs where everything has been specced up.

“The platform is a combination of five or six different models working in concert to ensure consistency of structure, interpretation, and colour, as well as the more general, generative image side of things. We've tailored these models to be more architectural focused.”

Proud says that while some architects have tried general purpose image generation platforms like Midjourney and Stable Diffusion, they tend to be only useful at the concept stage of design work where they just need a “vibe check” on their ideas.

“When they actually have to do something design specific, it's of no use to them, whereas they have already used our platform at multiple stages of the project alongside their processes, because it is reliable in doing what they need it to do.”

The story so far

Proud trained as an architect and spent time working in the design world in London. 

Just as the excitement about generative AI took hold in late 2022, he met his co-founder to be, Will Jones, an experienced software engineer with an interest in generative art and architecture.

“We were talking about all of the many myriad problems in architecture and design processes. And when ChatGPT, Midjourney, and Stable Diffusion came out, it seemed evident to us that if leveraged and tailored properly, [generative AI] would be game changing for architecture and design.”

Gendo’s product officially launched yesterday, following a two-month beta period, and an alpha test last year.

The startup names architecture firms like Zaha Hadid Architects, KPF, David Chipperfield Architects, and Benoy as having been among its beta test users.

Proud says beta testers generated almost 20,000 assets over the course of 10 weeks, with the software being used during a range of different stages of design work.

Gendo is now up and running as a browser-based SaaS subscription product that anyone can sign up to, with discounted pricing at launch.

Funding

Gendo has this week announced an £855,000 pre-seed round led by Concept Ventures, alongside Ascension Ventures, Baobab Ventures and angels with connections to AI and architecture, such as WeWork’s former Global Head of Design Studios.

Proud says the round was actually raised last summer. Many people don’t realise that it’s common practice to announce a raise long after it took place, for a range of reasons we’ll dig into another time.

Funds have so far been used to hire a small engineering team alongside some contracted assistance. The startup is currently hiring for a number of engineering, design, and marketing-related roles.

Now the first version of the product has launched, Proud says another round is likely “quite quickly” after they’ve assessed the initial impact this week’s launch makes.

Go deeper on Gendo

Subscribe to Premium to read the rest.