Browse
Log in / Register
Share
The automotive industry is undergoing a fundamental transformation, shifting its core definition from mechanical hardware to sophisticated software. At the forefront of this revolution is 42dot, a smart mobility solutions company under the Hyundai Motor Group, where developers are building the integrated systems that will power tomorrow's vehicles. On May 28, 2024, insights from key team leads reveal what it truly takes to create technology that moves beyond the screen and onto the road.
Gone are the days when a developer's job ended after writing code. The modern automotive software engineer must own the entire experience, from the algorithm to the actual driving behavior. This holistic mindset defines the teams at 42dot, where specialists in vehicle operating systems, AI agents, and autonomous driving collaborate to create a single, cohesive product.
The New Hiring Standard: Mindset Over Tools Sunghan Ryu, Team Lead for the Gleo Core AI agent, emphasizes that 42dot’s evaluation begins long before a candidate writes a single line of code. The critical question isn't just what technologies someone knows, but why they want to use them. "We look more at what you want to build with technology than at the technology itself," Ryu explains. This focus on purpose and vision ensures developers are aligned with creating tangible user experiences.
Junhee Lee, leading the Vehicle OS team, brings it back to fundamentals. Deep, persistent problem-solving ability trumps specific domain knowledge. His team includes former audio engineers and cloud database developers. "We value how deeply they understood the technologies they've used and how they pushed through difficult challenges," Lee states. This approach builds teams capable of innovating at the intersections of different tech fields.
Hayoung Kim of the autonomous-driving Trion team agrees. He seeks individuals who can independently define a problem and validate their solution from start to finish. Reflecting on his own interview, Kim notes it felt less like a search for correct answers and more like an exploration of his foundational thought process.
The Unbeatable Motivation: Seeing Code Drive Reality For many at 42dot, the work becomes personal when abstract algorithms manifest in physical movement. Hayoung Kim describes a pivotal moment: his team fed the same driving video to their AI model but changed only the navigation destination. The model correctly selected the appropriate lane for each turn. "Watching that, I felt our technology was actually leading the movement of a real vehicle," he shares.
Sunghan Ryu finds profound pride in knowing the software his team builds is operating on real roads. "As an AI developer, you often see projects stop at the Proof of Concept stage. Here, there's a real sense of reward knowing our efforts meet people through actual products," Ryu says. This connection from code to customer provides a powerful motivator absent in many tech sectors.
Junhee Lee draws energy from seeing how individual work contributes to the company's broader mission. Through internal tech events and forums, 42dot clearly communicates its technological roadmap. "In those moments, I can feel how an individual's achievement connects to the growth of the team and the entire organization," Lee observes.
Building a Culture for Diverse Talent With a mission to build software-defined mobility, 42dot attracts talent from vastly different backgrounds—a strength that requires intentional onboarding. The process focuses on contextual understanding and psychological safety. Ryu highlights the "Ask your Buddy!" program, which pairs newcomers with mentors to demystify workflows and accelerate integration.
Lee points to an ingrained culture where asking questions is not just accepted but encouraged. Teams often visit actual work sites or test vehicles together to give new members a direct, tactile understanding of development and validation processes. This collaborative spirit ensures learning continues on the job.
Kim adds that understanding the "why" is as crucial as reading the code. Teams meticulously document key technologies, experimental methods, and system architectures. They also hold regular sessions to share the team's strategic direction and current problem-solving focus, ensuring every developer sees how their task fits into the larger puzzle.
The Ultimate Responsibility: When AI Becomes an Experience For Ryu, joining 42dot instilled a new level of product responsibility. Previously, improving model performance might have been the primary goal. Now, the standard is consumer-grade reliability and safety within a moving vehicle. "We have to think about whether a feature works reliably in an actual car, how it should respond in unexpected situations, and how to handle user feedback after launch," he states.
This responsibility shapes daily work. The Gleo Core team uses a shared "Spec Wiki" documenting every function's operation, exception cases, decision logic, and testing standards. Even the rise of AI coding assistants hasn't lowered the bar. "The goal isn't just to write code faster," Ryu clarifies. "It's to develop and validate against a consistent standard across requirements, context, and tests to build a more reliable product."
This rigorous approach is vital for uncovering edge cases. A notable example occurred during "red teaming" for Gleo Guard, a safety module for the Gleo AI. Testers input a request using "dujjongku," Korean slang for a trendy cookie. The external AI API did not recognize this contemporary term, potentially misclassifying a simple navigation query as risky.
The team treated this not as a vocabulary bug but as a structural challenge. They progressed Gleo Guard into an in-house model, finely tuned for the automotive domain and capable of using such slang as contextual clues for intent. This proactive refinement, fueled by aggressive testing, highlights the exhaustive diligence required when AI directly controls a vehicle's functions.
As the industry accelerates, the demand for developers who bridge the gap between code and road experience will only grow. Companies like 42dot are setting a new benchmark, proving that the future of mobility depends as much on engineering mindset and product ownership as on pure technical skill. For developers like Sunghan Ryu, this evolution is personal. He smiles, sharing that his young daughter now points at cars and says "vroom-vroom." "I sometimes tell her that some of those cars may have technology I helped build," he says—a simple statement capturing the profound shift from writing algorithms to shaping journeys.
