Selected systems work2019 - PresentOngoing
Production AI Systems for Mobility
A confidentiality-safe snapshot of the kind of real-world AI systems work I have spent years doing: perception, in-cabin intelligence, optimization, and deployment under hard constraints.
What this represents
A large share of my career has been spent on systems that do not get the luxury of being merely interesting. They have to be performant, robust, observable, and carefully engineered around compute and safety constraints.
Publicly, I keep the description abstract enough to stay respectful of confidentiality while still showing the character of the work.
Work patterns
- Designing and optimizing multi-model, multi-sensor pipelines.
- Balancing model quality against latency, throughput, memory, and deployment limits.
- Working across architecture, execution planning, technical review, and customer-facing delivery.
- Building middleware and inference systems that support multiple perception components concurrently.
Why it matters to me
This work shaped how I think. It taught me to respect reality, to care about interfaces and constraints, and to distrust any system claim that only survives inside idealized conditions.