Builders Foundry London - Real World AI & Architecture

For people who actually build software. We focus on AI, software design, engineering, cloud and startups, without the hype.

No vendor pitches. No buzzwords. Just real architectures, production lessons and honest trade offs. For engineers, architects, platform teams and technical founders.

Join us for 2 practical talks and a roundtable conversation.

1. Using GenAI to Re-Imagine City Development
Damian Bemben - Senior Software Engineer @ AdaMode

"City re-development is a critical topic within infrastructure planning. Cities can be defined by the spaces built within them. Right now, opportuniteis for creative re-developments are limited - consultations and concept art is expensive & hard to produce. What if you could visualise a new development within a space in a few seconds? Damian will evaluate a genAI project, using a mixture of genAI techniques and latest image-to-3d models to examine the possibility of providing people with the opportunity to have a direct voice in urban planning. This talk will be a mixture between a technical explanation of effective model integration, and a wider discussion on GenAI for good; How do we find deeper, practical project success within models typically consigned to the "AI Slop" designation."

2. Deploying Deep Learning in Real Healthcare Constraints
David Agbolade - Senior Data Scientist, | AI Researcher | SheerFit Founder

"Here's the truth about deploying deep learning in healthcare: everything costs more and works less than you expect.
We built a system to generate radiology reports automatically because Africa has 1 radiologist per 1 million people. The model worked great in the lab: 0.347 BLEU-4, sub-60-second inference, published research. Then came deployment.
The actual challenges weren't the ones in papers. How do you collect 10,000 medical images when hospitals can barely afford IT? How do you handle bias when your dataset is 90% one condition? How do you deploy when the budget is $1,500 total? How do you convince doctors to trust predictions they can't explain?
This talk covers what actually worked: single GPU consumer hardware (no fancy cloud infrastructure), Streamlit for the prototype interface (because we needed something fast and simple), PyTorch with aggressive optimization (cutting everything that wasn't essential), and building attention visualisations doctors could actually interpret (not fancy academic stuff).
I'll share the uncomfortable decisions: shipping a Streamlit prototype because building a "proper" web app would take months we didn't have. Using a single model instead of an ensemble because inference time mattered more than 2% accuracy gains. Training on whatever data we could find rather than waiting for the "perfect" dataset that would never exist.
You'll see real architectures, actual costs, and honest trade-offs. No enterprise platforms, no vendor solutions, just what you can build with PyTorch, a GPU, and determination to ship something that works."

3. Round table
Tom Winstanley - CTO and Head of New Ventures in UK&I
Tom provides expert foresight into effective innovation and leading our startup and scaleup partnering programme, Bento Box. Tom has a background in consulting and programme delivery with over 20 years of international experience in telecoms, media, retail and financial services, always on the interface between business, customer experience and technology. Having co-founded the digital practice for NTT DATA in the UK, Tom remains active in client engagements, where he leads business, creative and technology teams, helping our clients rethink their businesses.

Open discussion on shipping AI systems under real world constraints.
Questions. Trade offs. What we would do differently.