AI-powered tools are reshaping product engineering, especially in industries where design speed and accuracy are critical. Simulation technology, generative AI, and intelligent platforms are evolving rapidly. This forces engineering firms to find new ways to accelerate development timelines while improving the quality and performance of complex products.
In a recent episode of the Analytics Insight podcast, host Priya Dialani spoke with Radha Krishnan, President and Founder of Detroit Engineered Products (DEP), about the real impact of AI on engineering teams. They discussed how AI-powered simulation and software development tools are filling critical gaps in the design process and output generation. The conversation delves deeper into how organizations think about workforce capacity. Here are the excerpts from the interview:
DEP, or Detroit Engineered Products, is a design engineering and software company headquartered in Troy, Michigan, with offices in India, China, and Japan, and partners across Europe. The company has grown to a team of over 550 engineers, the majority of whom hold advanced degrees in mechanical engineering, electrical engineering, and computer science.
DEP operates in the product design and development space, spanning automotive, aerospace, heavy equipment, and consumer goods. The company focuses on two areas. First is simulation-based engineering, which involves virtually modeling and validating product performance before hardware is built. Second is software development through the flagship Meshworks platform. Customers include global names like General Motors, Toyota, Maruti Suzuki, Ashok Leyland, and Volkswagen.
The company was founded after years at General Motors. This includes experience with the EV1, widely recognized as the first modern electric vehicle. Most lessons were shaped through real-world experience, building DEP from the ground up.
Today, the role focuses on setting the technology direction of the company, developing new engineering tools and processes. It also entails helping deploy these solutions to customers worldwide, all while staying hands-on with teams on a daily basis.
During the early stages of product development, designs evolve rapidly, sometimes multiple times in a single day. The key question is always whether a design will pass or fail its performance criteria. Traditionally, assessing something like vehicle crashworthiness required 20 engineers working for two months, by which point the design had already moved on.
With AI, models trained on historical simulation data can evaluate a new design against multiple performance criteria in a single day. It may not match the precision of physics-based simulation. However, it offers directional guidance that was simply not available before. This opens up an entirely new category of engineering work, and that means more engineers, not fewer.
Every software organization has far more work in its pipeline than it has capacity to deliver. If AI-assisted code generation allows a team to build 500 tools in the time it previously took to build 100. So, the answer is to build more, not reduce the team.
AI tools also shorten the time it takes for new developers to become productive. Hence, redirecting focus toward algorithm development and requirements definition, the genuinely creative parts of software engineering. The role evolves, but the need for engineers grows.
On the software side, AI-generated code acts as a strong starting point, even at 90% accuracy, that developers refine and validate through the full quality assurance cycle. This reduces the time to delivery without compromising standards.
On the simulation side, the key is training AI models on data that matches the specific product category. An AI model built on SUV data for a manufacturer that only makes SUVs will be far more accurate than a generic one. When applied correctly, AI bridges the gap between speed and accuracy in ways that were previously not possible.
Listen to the full discussion on the Analytics Insight Podcast.