Forest
Flager

The trajectory reads like a deliberate puzzle. Stanford for civil engineering. Harvard's Graduate School of Design for digital media. MIT for structural engineering. Back to Stanford for a PhD in computational design optimization. The connecting thread, across all of it, is the same question Forest has chased his entire career: how do you use computation to help humans make better decisions about how buildings are designed and built?

His 2003 SOM Foundation Structural Engineering Traveling Fellowship took him to Japan to examine sustainable development approaches - comparing traditional Japanese construction using renewable materials against Western design philosophy emphasizing structural robustness. The fellowship gave him an early lens on the relationship between material information and design outcomes.

That lens became a research career. His postdoctoral work at Stanford (2013-2017) focused on structural optimization of steel truss structures, building envelope design optimization, and multidisciplinary design optimization. He published extensively, built tools, and taught. Then Katerra recruited him as Director of Software and Design Automation in 2017.

Katerra was the SoftBank moonshot attempt to vertically integrate construction - controlling everything from design software to material supply to on-site assembly. Forest led the software team. When Katerra collapsed in 2021 after burning through over $2 billion, he had a sharper view than almost anyone of what the construction industry's real data problems were. Not design. Procurement.

Not everyone who worked at Katerra came out with a company. Forest did. The lesson he extracted from the collapse was specific: the construction industry's information problem is not a design problem. It is a procurement problem. Product data - specifications, pricing, availability, compliance - is fragmented across thousands of manufacturer websites, inconsistently formatted, manually maintained, and fundamentally unreliable at the moment it's needed most: during the bid.

Katerra had tried to solve construction by controlling the physical supply chain. Forest's instinct at Parspec was different: solve the data layer first. If you can reliably know what products are available, what they cost, what specs they meet, and what alternatives exist - then the workflow automation follows naturally.

This is why Parspec spent years building a product database before anyone was talking about AI in construction procurement. The 6 million products currently in Parspec's catalog didn't appear overnight. They were scraped, normalized, and maintained through a system of web crawlers and machine learning pipelines - the same basic approach Forest and Pratyush were using for academic research a decade earlier.

The insight that construction technology and domain expertise "rarely overlap" in the industry is the core of Parspec's competitive moat. Forest has both. His academic background gave him the computational fluency. His time at Arup and Katerra gave him the product knowledge. Co-founder Pratyush Havelia brings the data engineering depth. The combination is genuinely rare in the space.

Parspec's customer list - four of the five largest U.S. electrical distributors, collectively handling $70 billion in global annual sales - is the market's verdict on that combination. These are not venture-capital-friendly pilots. These are production deployments at companies that have been in business for decades and are notoriously slow to adopt new software.