The questions firms are actually asking.
These answers come from client discovery, AIA Los Angeles AI Office Hours, and the work of turning experiments into controlled workflows. Start with the question closest to your real work.
Start with the question you already have.
Which tool? Where should we start? What can go wrong? What will this cost? These answers begin with those questions, then explain the few decisions the firm needs to make around the work.
Start and fit
Where should an architecture firm start with AI?
Start with one recurring piece of real work—not a firm-wide tool rollout. Choose a workflow with a visible source, owner, review point, and destination, then learn what the team, data, and operating environment need before scaling.
Read the one-workflow starting guide →How advanced does our team need to be?
A firm can begin with one useful workflow and a team willing to examine how the work actually moves. Clearworks meets the firm where it is, builds practical fluency, and adds standards, administration, and governance as the work becomes more capable.
Find your firm-size path in the Playbook →Can we begin with one workflow?
Yes. A bounded workflow is often the best first move because the firm can see the inputs, handoffs, review requirements, ownership, and result clearly. The Blueprint is designed for that kind of focused start.
Explore the four ways in →Do we need an AI strategy first?
Start with the work. The useful strategy becomes clearer once the firm can see the real friction, the evidence behind it, the systems involved, and the decisions that matter. The strategy should emerge from operating reality rather than sit above it.
What makes a firm a good fit for Clearworks?
The strongest fit is a firm ready to expose a real workflow, assign an internal owner, involve the people who do the work, and participate in making the change stick. Clearworks is most useful when leadership wants an operating improvement, not simply another tool license.
Tools and implementation
Do we need to know which AI tool we want?
No. Start with the job the firm needs to improve. Clearworks compares whether to use what you have, buy, connect, or build based on the workflow, data, permissions, review requirements, cost, and who will own the result.
Read the tool-standardization guide →When should a firm use what it has, buy, connect, or build?
Use what you have when the existing system can support the workflow with clearer configuration or ownership. Buy when a mature product fits the job. Connect when the value is in moving trusted information between systems. Build when the workflow is distinctive enough to justify the maintenance.
Explore the workflow decision model →Can Clearworks work with our existing systems?
Yes. The first step is to understand what the current systems already hold, how people actually use them, where information breaks, and who owns each part. The useful fix may be configuration, integration, automation, a dashboard, training, or a new tool only where the evidence supports it.
How do you keep a pilot from becoming another abandoned tool?
Give the pilot a real owner, bounded use case, clear inputs, review gate, accepted destination, success measure, and scheduled decision point. A pilot should end with evidence about whether to adopt, revise, or retire the workflow—not an indefinite experiment.
Will this disrupt active projects?
The work is designed around live operations. Clearworks learns from what is already happening, bounds the first change, and keeps responsibilities and review points clear. Production changes should be introduced at a scale the firm can supervise safely.
Drawings, models, and delivery
How can architects use AI in rendering and visualization?
The most practical workflow starts from something real: a sketch, massing study, Revit model, SketchUp view, or existing render. AI can then help test materials, light, landscape, atmosphere, and presentation styles while the team keeps geometry and design intent anchored.
Read the model-grounded rendering guide →Can AI help create Revit views, sheets, tags, dimensions, or schedules?
AI-assisted tools can accelerate repeatable documentation work such as creating views, arranging sheets, placing tags and dimensions, and assembling schedules from a firm’s standards. The capability is promising but still requires bounded use, reliable source models, and professional review before issue.
See production workflows in the Atlas →Can AI review a drawing set for missing or inconsistent information?
Yes, in a review-assistance role. Current workflows can flag missing dimensions, inconsistent tags, unresolved selections, specification gaps, accessibility issues, or changes that did not propagate. The architect evaluates every finding and remains responsible for the issued work.
See drawing and QA workflows in the Atlas →How does AI help with markups, RFIs, submittals, and project handoffs?
The useful pattern is to turn a meeting, markup, site walk, RFI, submittal, field update, or change into a reviewed record with a clear next owner. Preserve the source, decision, and accepted update in the system where the project is managed.
See delivery and field workflows in the Atlas →Who approves AI-assisted production work?
The qualified professionals and project leaders already responsible for the work. AI can draft, compare, flag, organize, and assemble; architects and the appropriate consultants author, review, approve, and seal the deliverable.
Data, governance, and adoption
What firm data can an AI tool access?
That depends on the account, workspace settings, connected systems, permissions, retention terms, and the way the workflow is configured. Clearworks maps those boundaries explicitly so the firm can decide what information belongs in each tool and who can reach it.
What privacy, retention, and model-training settings should we review?
Review workspace ownership, user provisioning and offboarding, SSO where appropriate, retention, vendor model-training controls, connector permissions, approved tools, and the categories of project or business data each tool may receive. Settings should match the firm’s obligations and actual use.
Explore Embedded Advisory →What does practical AI governance look like inside an AEC firm?
Practical governance lives in the work: approved uses, data boundaries, review thresholds, ownership, workspace settings, training, documentation, and a clear path for evaluating new use cases. It is maintained as the tools and workflows change.
See how the program is run →Who should own the AI program?
One internal owner should coordinate decisions across leadership, operations, IT, security, legal counsel, and the people doing the work. The owner does not need to be the deepest technical expert; they need the authority and context to keep the program coherent.
See how ownership changes with firm size →How do you train people without giving another generic AI demo?
Train by role and real task. People should practice with the work they recognize, learn what a good input and review look like, understand the data boundary, and leave with a reusable workflow or playbook. Follow-up reveals what actually stuck.
Explore workshops →Engagements
What is a Blueprint?
A Blueprint is a short audit of one documented workflow. It clarifies the operating model, exposes the friction and ownership, and leaves the firm with a practical first fix rather than a broad transformation plan.
What happens during a Full Busywork Audit?
Clearworks traces how work actually moves across the firm, inventories the systems and evidence, identifies operating and adoption constraints, quantifies opportunities where the evidence supports them, and produces a prioritized roadmap leaders and staff can evaluate together.
What happens after the audit?
The firm can use the roadmap internally. Clearworks can stay involved through Embedded Advisory, facilitate training, or scope a specific implementation when the evidence shows one is warranted.
Explore Embedded Advisory →What does Embedded Advisory include?
Embedded Advisory keeps the AI operating environment and roadmap moving. It can include administration, role-based training, practical governance, adoption support, workflow improvement, and the next implementation worth pursuing.
See the full advisory offer →What happens in a Clearworks workshop?
A workshop builds practical fluency around the firm’s own work. Participants practice a bounded workflow, learn the relevant review and data boundaries, and leave with reusable examples or playbooks plus a follow-up check on what the team adopted.
See workshops →