Why AI matters at this scale

CDM Smith is a large engineering and consulting firm specializing in civil engineering and infrastructure projects. With a workforce of 5,000 to 10,000 employees and operations dating back to 1947, the company manages complex, long-term projects such as water systems, transportation networks, and environmental facilities. At this scale, even small efficiency gains translate into significant financial and competitive advantages. The engineering sector is undergoing a digital transformation, and AI is a pivotal technology for maintaining leadership, improving project outcomes, and addressing growing infrastructure demands.

For a firm of this size and maturity, AI adoption is not just about innovation but operational necessity. Large project portfolios generate vast amounts of data—from design documents and sensor feeds to historical performance metrics. Leveraging AI can unlock insights from this data, moving from reactive problem-solving to predictive and optimized project delivery. The size band allows for dedicated AI pilot teams and investment, but also introduces challenges in change management across a geographically dispersed, experienced workforce.

Concrete AI Opportunities with ROI Framing

1. Generative Design Automation: Implementing AI-driven generative design tools can automate the creation of multiple infrastructure design alternatives. By inputting constraints like materials, costs, safety standards, and environmental impact, the AI proposes optimized designs. This reduces manual drafting time by an estimated 30-50%, accelerates client presentations, and can lead to more sustainable, cost-effective solutions. The ROI comes from shorter project cycles and reduced labor costs on repetitive design tasks.

2. Predictive Project Analytics: Machine learning models can analyze decades of historical project data to identify patterns leading to delays, budget overruns, or safety incidents. By flagging high-risk projects early, managers can allocate resources proactively. For a company managing hundreds of projects annually, reducing average overruns by even 5-10% through better prediction could save tens of millions of dollars, providing a clear and rapid ROI.

3. Automated Compliance and Document Management: Natural Language Processing (NLP) can review thousands of pages of technical specifications, regulatory documents, and permit applications to ensure compliance. This reduces the risk of costly errors or violations and frees senior engineers from tedious review work. Automating this process could cut compliance review time by up to 70%, improving margins on fixed-fee projects and enhancing reputation for reliability.

Deployment Risks Specific to This Size Band

Deploying AI at a large, established firm like CDM Smith carries specific risks. Integration Complexity: Legacy software systems for CAD, project management, and finance may not easily connect with modern AI platforms, requiring costly middleware or replacement. Data Silos: Information is often trapped within individual project teams or regional offices, making it difficult to aggregate the high-quality, unified datasets needed for effective AI. Cultural Inertia: With a long history and standardized practices, engineers and managers may be resistant to adopting AI-driven workflows, perceiving them as a threat to expertise or requiring significant retraining. Scalability vs. Pilot Purgatory: While the company has the resources to fund pilot projects, there is a risk of these initiatives remaining isolated "science experiments" without a clear strategy for organization-wide scaling, diluting potential ROI. Mitigating these risks requires strong executive sponsorship, phased integration plans, and focused change management programs that demonstrate AI's value as an enhancer of human skill, not a replacement.