Senior Software Engineer (Embedded Ai)
- Primarily concerned with the development of high performance firmware to enhance inference speed of computer vision algorithms on TI-SOCs making use of on chip DSPs and optimized DMA engines.
- Developed a generic, scalable dataflow template to optimize memory usage during model execution on chip. Additionally contributed to and improved the off-chip memory manager to make better use of on-chip memory.
- Worked with deep learning architectures (CNNs, transformers) for automotive applications, including in-cabin monitoring and object detection using multimodal sensor data.
- Scaled the firmware to work across a wide range of devices with varying memory arrangements and compute units. Increased device coverage by 4x.
- Worked with customers to fine tune models for efficiently working on supported devices as well as close reported bugs on our E2E forums.
- Led a team to improve functional safety of firmware to achieve ASIL-B level compliance.