Hi there,

This week's three posts started in different places and ended up in the same one. The first argues that Revit is the wrong tool for the first few weeks of a project, and that the value of an AI-augmented early-stage workflow sits in the links between the tools rather than in the tools themselves. The second comes from a roundtable I co-chaired with Revizto in London the week before last, where heads of design technology, BIM and AI from some of the largest international practices were candid about which AI tools they would actually keep and the mistakes they are seeing across the industry. The third looks at bids and proposals, where most practices still move between SketchUp, Excel, CAD, Illustrator and InDesign with nothing connecting them. The common thread is one I keep returning to: the problem is rarely the tools, it is the integration between them.

Below you will find the three posts, a reminder of our free From Data to Decisions paper with Epic Games and the webinar that follows it on 21 October, where to find me at RELEASE [AEC] in Paris, a short survey from The Outsource Studio, and five new tools in the ADDD Database.

Let's get into it. 👇

The ADDD Workshop: a practical starting point for AI and technology decisions

The Workshop is a focused half-day session, held in person or online with your team, that looks at where technology and AI can help your practice and where they are currently causing friction. We begin by talking through how your team actually works, from early design through to delivery, so that the conversation is grounded in your real workflows rather than in a generic view of what a practice should be doing. From there we go through practical examples of tools that are relevant to your discipline, which gives everyone a shared and realistic picture of what is available and what is worth paying attention to.

I have been running workshops for Orms, Haworth Tompkins, Levitt Bernstein, DMWR, EPR and Morris & Co.

Within the session we identify the main problems holding the practice back, agree on which opportunities are worth pursuing first, and talk openly about the cost of leaving things as they are. Within a few days you receive a written report that sets out the key findings, the priorities and the recommended next steps, so the discussion turns into something you can act on and share with the wider team or the partners.

The Workshop is intended to give you strategic clarity, and it is not a technical audit of your systems. It works well as a first step for practices that know their tools and processes need attention but are not yet sure where to begin, and it can lead on to deeper work such as a early stage workflow software, Technology Blueprint or ongoing advisory support if that turns out to be useful.

Book a call here: 30 min with Allister Lewis

The Outsource Studio is running a short survey on how studios and independent designers in the built environment find talent and work, and where platforms, recruiters and outsourcing companies fall short. It takes about five minutes, and everyone who leaves their email receives the results.

Take the survey here: The Outsource Studio survey

YouTube

Instagram

The Revit post is the one I would ask practice leaders to sit with this week. Most practices moved early-stage design into Revit for a sensible reason, keeping everything in one integrated model, but the cost is that decisions get forced before the team is ready to make them, iteration slows to the pace of the geometry, and the metrics coming back at this stage are basic.

The roundtable explains why this matters beyond one stage of work. When senior digital leads were asked which AI tools they would keep, they named platforms that let them switch between models and only tools they had tested against a defined purpose, and when they were asked about the biggest mistakes, they pointed to AI applied to workflows that were already broken and saved time that disappears into more options and more content.

The bid workflow makes the same case for proposals: the value is in a chain that holds together, so one change updates everything downstream, and this week's new tools follow the same pattern.

The practical step is to look at the first few weeks of your next project before anything is opened in Revit. Write down which decisions the team actually needs to make at that point, which tools they are using to make them, and where information gets retyped, exported or redrawn between those tools. That map tells you where an AI-augmented workflow would speed up iteration and where it would only speed up a process that is already broken. This is the early-stage design workflow work I have been doing with practices: mapping how the opening stages of a project really run, then building and testing a connected workflow around them in a controlled environment, with people in the loop, before it goes anywhere near a live project. If you would like to explore what that could look like in your practice, the easiest way to start is a 30-minute call.