This article provides an executive summary of our recent webinar, “AI in Architectural Practice: Real Projects, Real Impact”, delivered by CMap in partnership with Architects’ Journal. The session brought together leading voices from architecture, technology, automation, and practice management to explore how AI is already transforming the way firms operate, from design development and client collaboration to housing delivery and business intelligence.

AI in architectural practice: Real projects, real impact

Artificial intelligence is reshaping the architectural profession, but not through the dystopian narrative of replacing architects. Instead, as this webinar made clear, AI is quietly revolutionizing workflows, unlocking capacity, reducing risks, and enabling architects to focus more on creativity and project quality. Hosted by Kunle Barker, the session brought together four leaders using AI in tangible, high‑value ways across practice management, housing delivery, automation and robotics, and feasibility analysis.

The panel featured:

• Dan Stanton, Associate Director, BDG Architecture + Design
• Mollie Claypool, CEO & Co-founder, Automated Architecture (AUAR)
• Russell Curtis, Director, RCKa Architects
• Ben Jervis, Product Director, CMap

Together, they painted a clear picture: AI isn’t the future of architecture - it’s already embedded in real projects delivering measurable results.

‍

Using AI to “carry your client in your pocket” - BDG’s virtual client agents

Dan Stanton opened with BDG’s approach to prioritising purpose over shiny tools. Rather than adopting every new model or AI solution, BDG first asks: What real problems do we need AI to solve?

For BDG, the challenge is scale - designing for major global brands like Google, LEGO, and Sony means handling thousands of pages of guidelines, audits, standards, and briefing documentation.

This is where BDG’s virtual client agents come in.

Imagine being able to ask a detailed question, test an idea, or cross‑reference a specific piece of briefing information instantly rather than having to wait for your next meeting or client engagement session.

Using Retrieval-Augmented Generation (RAG), BDG ensures that any answer the model gives is sourced directly from verified project documentation - removing hallucinations and drastically reducing the time teams spend hunting for information.

Most importantly:

• Junior staff now have instant access to client knowledge
• The design team stays fully aligned with the brief
• Gaps or inconsistencies in client documentation become visible early

Dan summarised BDG’s approach simply: Less time searching. More time designing.

‍

Automated, deployable micro-factories for housing - AUAR’s “physical AI”

Automated Architecture (AUAR), led by CEO Mollie Claypool, is tackling two huge societal problems simultaneously: the housing crisis and the climate crisis.

Their solution: deployable micro-factories that can manufacture timber housing cores and shells directly on, or near, construction sites. Mollie explained:

‍The builders we work with don’t buy robots. They buy output.

Rather than investing in high‑risk fixed factories, AUAR operates compact, containerised micro-factories powered by its “Master Builder” physical‑AI software.

This system:

• Converts a developer’s existing designs into production‑ready instructions within hours
• Handles variability in real materials (like knots or warping in timber)
• Continuously learns from each deployment
• Reduces onsite waste and speeds up delivery

The result is a scalable, low‑risk way to increase housing supply while improving quality and sustainability.

‍

AI to unlock housing capacity at scale - RCKa’s geospatial AI for small sites

Russell Curtis showed a dramatically different, yet equally powerful, use of AI: training convolutional neural networks to identify and classify developable small sites across London.

The initial challenge was simple: London’s small‑site housing capacity was vastly underestimated. Using AI, RCKa assessed 2.1 million freeholds, classifying and mapping potential sites that traditional methods had missed.

We reckon at the moment there’s probably space for around 800,000 homes on small sites across the whole of the city.

This evidence is now informing:

• Local authority small-site strategies
• Consultancy for planning departments
• The Greater London Authority’s new small-site design code
• Policy thinking around suburban intensification

Russell emphasised that none of this analysis would have been feasible without AI’s scale and pattern‑recognition capabilities.

‍

Using AI to transform practice management - CMap intelligence

Instead of exploring AI in design or construction, CMap’s Product Director, Ben Jervis focused on an equally critical but often invisible layer: the daily operational friction that slows architectural practices down.

Many firms still wrestle with time‑consuming admin - resourcing decisions, fee modelling, chasing data across spreadsheets, and the never‑ending task of managing emails. These essential activities keep a business functioning, but they can also drain time and energy from the creative work architects are trained for.

CMap intelligence was built to change that. By combining AI, automation, and a unified data layer across CRM, finance, project deliver & information, the platform acts as a practice-wide intelligence layer. Its suite of agents - covering fee estimating, resourcing, email filing, and real‑time insights - helps practices operate with greater clarity, consistency, and confidence.

As Ben put it, the aim is simple: free people from manual tasks so they can focus on higher‑value work.

When you get AI right behind the scenes, you’re effectively winning the time back to do the things you became an architect for in the first place.

‍

‍Common themes across the discussion

AI solves problems - it doesn’t replace your architects

A strong consensus emerged across the panel: AI’s real value lies in removing friction, not removing people.

The panel emphasised that AI is best suited to handling repeatable, time‑consuming, or information‑dense tasks that pull architects away from design thinking. Whether it’s searching through hundreds of pages of client standards, running viability checks across thousands of sites, or sorting emails that no one has time to file, AI is becoming an intelligent assistant, one that accelerates the work without undermining architectural expertise.

What it frees, in return, is time: time to design, time to think, time to engage more creatively with clients & communities.

‍Start with a problem, not a tool

A recurring message was the importance of resisting “solution-first thinking.”

With new AI tools emerging almost daily, it’s easy to feel compelled to adopt something simply because it exists. But the panel agreed that meaningful impact comes from identifying a specific friction point - an inefficiency, a blind spot, a slow process - and then selecting or shaping an AI approach to solve that problem.

The firms seeing real benefits aren’t chasing models or acronyms; they’re grounding their experimentation in real workflow challenges, then iterating from there. AI becomes effective not because it is sophisticated, but because it is purposeful.

‍AI increases access and equity inside teams

Another powerful theme was the way AI can democratise knowledge within a practice.

Traditionally, valuable project or client knowledge sits with senior team members, accumulated through years of experience or direct client interaction. Tools like BDG’s virtual client agents flip that hierarchy - suddenly, junior designers can ask detailed questions, access verified information instantly, and gain clarity that would previously require waiting for a meeting or tapping a senior colleague on the shoulder.

This creates more confident teams, smoother workflows, and a more level playing field for emerging talent. Instead of replacing people, AI can elevate them, especially those earlier in their career.

The human role - judgment, trust, design vision - remains irreplaceable

Despite the sophistication of the tools discussed, every panellist returned to a central truth: architecture is a human profession.

Clients don’t buy algorithms; they buy judgment, trust, interpretation, taste, and vision. AI can support decision‑making, structure information, validate assumptions, or surface risks, but it cannot take responsibility for a design outcome or relationship. As Dan summarised:

You’re still going to have to sit in front of a client and make a judgement call.

AI may provide insight, but the architect provides direction. The panel agreed that this combination, rather than competition, represents the future of practice: architects supported, not replaced, by intelligent technology.

‍

Final thoughts

This session made one thing clear: architecture is entering an era where AI isn’t optional, it’s a competitive advantage. Whether through robotics, data intelligence, urban planning, or business operations, firms that adopt AI meaningfully will free their teams to focus on creativity, design quality, and strategic decision‑making.

CMap is proud to support architects on that journey, helping practices unlock time, insight, and confidence as AI becomes an everyday part of professional life