Week 7 - Learning Journal

 For this weeks learning journal:

I've interviewed an industry expert at Qualcomm.

Industry Expert Interview – Learning Journal Entry

Khiem Vo  |  3rd Year, Computer Science  |  CSUMB

Introduction

For this assignment, I chose to interview Joel Gorin, a Software Engineer at Qualcomm. Qualcomm is one of the most influential semiconductor and technology companies in the world, known for pioneering mobile chipset design, wireless technology, and increasingly, on-device AI and machine learning solutions. As a third-year Computer Science student with a strong interest in AI and machine learning, I was drawn to Joel because his work sits at an intersection I find deeply compelling — low-level systems engineering that enables high-level intelligence. Understanding how AI is being implemented at the hardware and firmware level, not just in cloud-based software, felt like a perspective I could not get from a textbook.

I reached out to Joel through LinkedIn, introducing myself as a CSUMB student interested in learning more about how large technology companies approach software engineering in the context of emerging AI capabilities. He graciously agreed to a 25-minute video call, and the conversation turned out to be one of the more valuable I have had in my academic journey so far.

Summary of Key Takeaways

Joel's career path was not a straight line, which was one of the first things that stood out to me. He studied Computer Science and moved through roles of increasing technical complexity before landing at Qualcomm, where he works on software that interfaces closely with hardware — a domain that requires both systems-level thinking and an understanding of how end-user applications will eventually consume the capabilities being built. He emphasized that being adaptable and willing to go deep on unfamiliar topics was more important than always having the right background on day one.

When I asked about the most significant challenges in his role, Joel pointed to the growing complexity of debugging in systems where software, firmware, and hardware are tightly coupled. When something goes wrong, isolating the cause requires a level of patience and methodical reasoning that, he noted, many junior engineers underestimate. He also mentioned that cross-functional communication — being able to explain deeply technical problems to non-engineers or to engineers from different disciplines — is a daily challenge and a critical skill.

On the topic of emerging trends, the conversation naturally gravitated toward on-device AI. Joel spoke about how Qualcomm is investing heavily in enabling machine learning inference to run directly on mobile and edge devices rather than relying on cloud computation. This has significant implications for latency, privacy, and energy efficiency. He described this shift as one of the most exciting and demanding frontiers in the industry right now — one where software engineers need to understand not just algorithms but the computational constraints of the hardware they are targeting. He mentioned tools like Qualcomm's AI Engine and the growing importance of model optimization techniques such as quantization and pruning, areas I had touched on in coursework but now see with much more real-world urgency.

When I asked what skills he values most, Joel's answer surprised me a little. He said strong fundamentals — data structures, algorithms, operating systems — matter far more than fluency in any particular framework or language. Frameworks change; foundational thinking does not. He also stressed the importance of curiosity and ownership: engineers who ask good questions and take responsibility for their work, even when it is uncomfortable, are the ones who grow fastest.

His advice to students was direct: build things. Do not wait until you feel ready. Work on personal projects, contribute to open source, and get comfortable being a beginner repeatedly, because the industry moves fast enough that everyone is a beginner again every few years.

Reflection

Sitting with this conversation afterward, I found myself thinking about the gap between how I had been approaching my education and what Joel described as the reality of working in industry. I have spent a lot of time in my coursework focused on getting things to work — passing tests, meeting assignment requirements — but Joel's emphasis on deeply understanding why something works (or does not) gave me pause. The kind of debugging and systems thinking he described is not something you can cram for. It has to be built slowly, through practice and genuine curiosity.

What struck me most was his framing of on-device AI not as a niche specialty, but as a growing mainstream expectation in the industry. I had been thinking about AI and ML largely in terms of training large models or working on cloud-based applications. But the reality Joel painted is more nuanced — a world where understanding how to make models efficient enough to run on constrained hardware is just as important as building the models themselves. That reframing felt significant. It made me want to pay closer attention to topics like systems programming and hardware architecture, which I had mentally filed away as less relevant to AI. Clearly, they are not.

I also appreciated his honesty about the unglamorous parts of the job. There is a tendency in tech culture to highlight the exciting, cutting-edge work and gloss over the hours spent reading documentation, tracking down obscure bugs, or rewriting code that almost works. Joel did not romanticize the job, and I found that more motivating than any polished pitch would have been. It made the career feel real and achievable.

Future Steps

This conversation gave me several concrete directions to pursue. First, I plan to spend more time on model optimization — specifically learning about quantization and how inference pipelines are built for edge deployment. I want to go beyond the high-level conceptual understanding I have now and get hands-on experience with tools like ONNX or TensorFlow Lite. Second, I am going to be more intentional about strengthening my systems fundamentals. Joel's comment about foundational knowledge resonated with me, and I want to make sure my understanding of operating systems and memory management is solid before I graduate. Third, I am going to start a small personal project — something that combines an ML component with a systems constraint — so I can begin building a portfolio that reflects the intersection of AI and hardware-aware software. Finally, I would like to stay in touch with Joel and continue building my professional network with engineers who work at the boundary of hardware and software, since that is increasingly where I see myself wanting to work.

This interview reminded me that career growth is not just about accumulating skills — it is about developing judgment, curiosity, and the resilience to keep learning even when it is hard. I am grateful for Joel's time and candor, and I am leaving this experience with a clearer sense of the kind of engineer I want to become.