AI Agent Operational Lift for Emh&t in Columbus, Ohio

Why AI matters at this scale

emh&t is a century-old civil engineering firm headquartered in Columbus, Ohio, specializing in transportation and infrastructure projects. With 201-500 employees, the firm operates at a scale where it has the project volume to generate meaningful training data but lacks the vast IT budgets of global engineering conglomerates. This mid-market position creates a unique AI opportunity: the firm can be more agile than larger competitors while having enough resources to deploy targeted, high-ROI solutions.

Civil engineering has historically been a slow adopter of AI, relying heavily on manual CAD work, experience-based judgment, and paper-driven processes. However, the sector is reaching a tipping point. Labor shortages, tightening project margins, and the increasing complexity of infrastructure funding (such as the IIJA) are forcing firms to seek efficiency gains. For emh&t, AI represents a way to differentiate its bids, reduce overhead, and attract younger talent who expect modern tools.

Three concrete AI opportunities with ROI framing

1. Automated proposal and bid management. Public infrastructure projects require extensive RFP responses, often running hundreds of pages. An NLP-driven system can ingest RFPs, extract requirements, and draft compliance matrices and technical narratives. For a firm submitting 50+ proposals annually, saving 100 hours per bid at a blended rate of $150/hour yields $750,000 in annual savings, while potentially improving win rates through faster, more consistent responses.

2. Generative design for preliminary engineering. Roadway and intersection design involves evaluating countless geometric alternatives against cost, environmental, and safety criteria. Generative AI tools can produce and rank thousands of options in hours rather than weeks. On a typical $5 million design contract, reducing preliminary engineering effort by 30% frees up $150,000 in billable capacity that can be redirected to additional projects or scope.

3. Predictive maintenance for asset management contracts. State DOTs are increasingly outsourcing long-term asset management. By training models on inspection records, traffic data, and material degradation curves, emh&t can offer condition-based maintenance schedules that reduce client lifecycle costs by 15-20%. This transforms the firm from a reactive design shop into a strategic infrastructure advisor, opening recurring revenue streams.

Deployment risks specific to this size band

Mid-market firms face distinct risks. First, data fragmentation is acute—project files live on individual engineers' machines, shared drives, and legacy document systems. Without a centralized data lake, AI models will underperform. Second, change management in a 100-year-old firm can be challenging; senior engineers may distrust black-box recommendations. A phased approach with transparent, explainable AI outputs is essential. Third, cybersecurity and IP protection must be addressed when using cloud-based AI tools on sensitive infrastructure designs. Finally, the firm must carefully manage professional liability—any AI-assisted design must still be sealed by a licensed Professional Engineer, requiring clear human-in-the-loop workflows and audit trails.