Taking the Engineering AI pulse from Silicon Valley to Detroit

Adopting Engineering Intelligence requires alignment at every level of an organization, from the engineers working with AI tools daily to the executives whose strategy depends on it. In one week, conversations about building AI into engineering happened at every level of the industry: A Customer Advisory Board in Silicon Valley, an executive dinner co-hosted with Goldman Sachs in Detroit and the first hands-on bootcamp of 2026 at AWS's offices, where engineers built production-grade AI workflows instead of watching demos. Three formats, one consistent conversation: how fast AI is moving into engineering production, what the real challenges are, and where to take this next.

‍

Executive Field Notes

The Bay Area kicked off the week with a Customer Advisory Board, where technology executives dug into the strategic implications of AI-driven engineering at scale. Companies in the room were the ones pushing the platform hardest, where product direction, deployment challenges and real-world edge cases were worked through together. Some candid conversations that rarely make it into a deck. 

Detroit took this further: Neural Concept and Goldman Sachs co-hosted an executive dinner bringing together senior leaders from automotive suppliers, OEMs, and technology partners. The evening centered on a shift the room knew well: Engineering Intelligence in automotive is no longer an implementation question. The industry is at the scaling stage, and the harder questions are organizational. 

‍‍

‍"This is a true change management effort that goes well beyond the technology. Evenings like yesterday really contribute to pushing our sector forward."

— Senior executive, major automotive supplier

‍

Across both regions, three themes kept surfacing:

1. Engineering Intelligence is moving into production faster than most expected.

The gap between research and the factory floor is closing. Fast Company recently reported that major automakers like General Motors and Jaguar Land Rover are already running hundreds to thousands of aerodynamic analyses per day with AI — a workflow that didn't exist two years ago. Instead of running isolated pilots, companies are embedding AI into core design workflows, with teams reporting 30–70% shorter iteration loops and real-time predictions that would have required hours to compute just a couple of years ago. What felt like a 5-year horizon is in fact happening now, and companies still in evaluation mode have less runway than they think. Early movers have accumulated months of workflow adaptation and embedded know-how that doesn’t transfer overnight. 

‍

2. Encoding engineering expertise directly into AI-driven workflows is becoming a structural competitive advantage.

The teams seeing the biggest gains have moved past tool adoption. They encode years of engineering expertise directly into the platform, training it to reflect how they think. One engineer can now cover design territory that used to require an entire team. As one customer put it:

‍"A single new design would take roughly 2.5 weeks. With Neural Concept, we can evaluate 13 new designs in around 30 minutes."

With that kind of throughput, the shape of engineering work starts to change. The platform becomes a representation of how a team thinks about product design: the tradeoffs they prioritize and the constraints they've learned to navigate. That's a competitive asset that grows with every project.

‍

3. Scaling across regions, departments, and governance structures is the next frontier. 

Early adopters are already thinking about what comes next: expanding from 20 engineers on the platform to 100 or 200, across geographies and business units. The question is organizational: how do you replicate the impact, maintain governance, and bring along teams who haven't yet experienced AI-native engineering? No standard playbook exists yet, but a pattern is emerging. The organizations making real progress are treating AI deployment as an engineering discipline in itself: starting with one high-impact workflow, building internal expertise around it, then expanding from there. Organisations that stall are usually doing one of two things: trying to transform everything at once…or waiting for the perfect moment to start.

‍

From Boardroom to Bootcamp: Engineering Field Notes

The next morning, the conversation moved from the dinner table to the workshop floor.

Neural Concept's first bootcamp of 2026, hosted by AWS in Detroit, brought together 30+ engineers for a full day working on the platform. The topic: F1 aerodynamics. Working through AI-driven geometry generation and aerodynamic design space exploration, participants moved from a single starting point to thousands of evaluated design variants, with performance feedback across physics domains running in real time.

Engineers who arrived focused on a single design problem left thinking about an entire design space, and certified to lead that shift in their own engineering teams.

‍

Engineering Intelligence as Industrial Infrastructure

Industrial manufacturing is navigating a genuinely difficult moment. Slowing sales, shifting trade policy, new competitive entrants, product cycles shortening while complexity increases.

From the conversations happening at both the executive and engineering levels, one thing is becoming clear: the technology question has largely been answered. The tools exist, the results are documented, and the deployments are real. The harder question now is organizational — getting from a proof of concept with one team to a standard way of working across an entire company.

The gap between the teams who've worked this out and those who haven't yet is widening. And it’s widening fast.