Muhammad Sufian

Embedded ML Engineer

SovIoTech

Oct 2024 - Present • 1 yr 7 mos

As an Embedded ML Engineer at Soviotech for over a year, I specialize in bridging the gap between resource-constrained hardware and intelligent software. My focus lies in architecting high-performance firmware and integrating AI capabilities into edge devices, utilizing platforms ranging from the STM32U5 series to the ESP32-S3 and Raspberry Pi. My core technical achievements include: Edge AI & Wearables: Leading firmware development for AI-integrated smart glasses, utilizing the LiveKit SDK on ESP32-S3 for real-time audio processing and low-latency inference. Hardware Acceleration: Spearheading the migration of legacy display drivers from QSPI to hardware-accelerated OCTOSPI interfaces on STM32 architectures, significantly optimizing throughput and rendering performance. Control Systems: Designing robust flight control algorithms and PID tuning for UAVs, ensuring stability and fault tolerance in dynamic environments. Embedded Linux: Engineering custom Debian-based firmware installers and multimedia drivers for Raspberry Pi integrated systems. I am passionate about pushing the limits of TinyML and building secure, efficient embedded ecosystems that solve real-world problems.