Partha Singha Roy

Embedded Systems Engineer specializing in C/C++ and Embedded Linux for ARM-based Edge AI systems. Hands-on experience with Yocto, Linux kernel, device drivers, V4L2, DMA-BUF, and Rockchip MPP. Experienced in real-time video pipelines, CNN-based object detection, and AI acceleration across Rockchip and Renesas platforms.

Embedded Software Engineer at Naveera Tech. (2026-06 – Present)

• Developing an AI-enabled dashcam and collision-detection system in C/C++ on the Rockchip RK3576 platform, leveraging its 6 TOPS NPU for real-time on-device AI inference and object detection.
• Implementing a dual-camera video pipeline using IMX415 and OV5647 sensors with V4L2 capture, DMA-BUF buffer sharing, H.264 encoding, and 60 FPS recording.
• Designing concurrent software components using Inter-Process Communication (IPC) and POSIX threads for camera capture, video encoding, AI processing, and event handling.
• Developing high-performance, low-latency video processing components using V4L2, DMA-BUF, and Rockchip MPP for efficient zero-copy data handling.

Embedded System Engineer at Aarish Technologies Inc. (2024-07 – 2026-06)

• Built a C++ multi-threaded POSIX application for audio CNN-based Battlefield Surveillance Radar sound classification and implemented logarithmic and linear AGC algorithms.
• Engineered a custom SDR communication system on PlutoPlus for real-time H.264 video and drone telemetry transmission using OFDM-QPSK baseband processing with liquid-dsp; implemented wideband frequency hopping across 900 MHz–5.5 GHz and achieved data rates up to 735 kbps.
• Created a C/C++ smart vision system on Raspberry Pi and ESP32-C3 using YOLOv3, RTP streaming, GPS/PIR/MEMS sensors, and ping-pong frame buffering.
• Developed a modular C++ framework for Renesas RZ/V2L to execute CNN models on the DRP-AI accelerator; built a Yocto-based Embedded Linux platform with kernel customization, DTS, and board-specific layers.
• Integrated IT6161 HDMI-MIPI bridge, IMX219/OV5647 camera sensors, and Wi-Fi drivers in Linux kernel 5.10; led RZ/V2L SBC design and production, reducing BOM cost by 40%.

Bachelor of Technology (B.Tech.) in Electronics & Communication Engineering – Kalyani Govt. Engineering College (2020-08 – 2024-07)