Overview
Alina Holovatiuk — an AI-in-architecture enthusiast and founder of the musical startup InTempo, based in Kyiv — received an Honourable Mention for a proposal that lets Old Spitalfields Market change use through the lightest possible means: fabric, lighting, colour and zoning, rather than heavy construction. Her entry asks how the existing iron structure can be amplified rather than obscured, and how the same space can host very different events across a day, a season or a circumstance.
The concept
The building's iron structure stays the protagonist. Around it, Alina proposes a kit of simple devices — fabric, lighting and colour — that divide the space into zones and let the market shift between markets, performances, exhibitions and civic gatherings with minimal effort. Wayfinding, advertising and event information are projected onto the fabric or hung at higher levels to avoid visual chaos at human height. The colour of the fabric, and the way it is hung, becomes a visual signal of the type of event happening at any given moment. It is a proposal about choreography as much as architecture: a system that can change quickly, equally well enjoyed by people of all ages and nationalities.
Workflow & process
Alina came to the competition late and worked at speed, using Gendo to move from nothing to coherent concept in a very short window. She rarely starts with hand sketches, preferring to manage volumes in 3D from the outset, and used Gendo to test the spatial logic of the fabric zones, the projected wayfinding and the changing colour palette. The people-cutout generation and populate function were standouts for her, letting the scenes feel inhabited without slowing the iteration loop. For next time she plans to display more of the thinking process — schemes and intent — directly on the board, so the reasoning behind the variations is as visible as the variations themselves.
Interview
Q: Was this your first time using Gendo, and how has your relationship with it evolved?
First time using Gendo. Very user-friendly — even without long tutorials. I was curious about the models the agent picks for different tasks, since output quality varies. The people-cutout generation and populate function were among the best parts.
Q: What did the result mean to you — and what would you do differently next time?
Next time I'll add more of the thinking process — the schemes that were in my head — onto the boards. And I'll start earlier, because I didn't have enough time to show the work on the entrances. Since this was the first AI contest I entered, I now know how much time each part of the process takes.
Q: In your own words, what was it like to compete in Gendo 48?
Gendo 48 was about rebooting how concepts are created with AI, and the pure joy of doing this. It was a test of how quickly the concept and its variations could be completed, and of how to stay reasonable to the task and grounded in the 'infinite' AI possibilities.
Key takeaways
Alina's proposal makes its case through lightness. By choosing fabric, light and colour over construction, the project keeps the iron hall legible and lets the building's civic function flex on a daily basis. It is also a clear-eyed first-time-user account: a reminder that Gendo can be picked up quickly, and that the discipline of showing one's thinking — not just one's images — is what separates good entries from great ones.