Rohan Kumar Singh

Entry-level Embedded Systems Engineer with hands-on experience building bare-metal firmware on ARM Cortex-M4 microcontrollers (STM32 F4 series). Delivered a research-grade, real-time Parkinson's disease speech detection system on STM32F401RE — the first known implementation of a full MFCC and NMF pipeline as bare-metal firmware on a Cortex-M4, with on-chip NanoEdge AI inference and zero host-PC dependency. Currently developing a real-time dysarthria detection system using GFCC, NNMF, and Graph Signal Processing on STM32F407VET6.

Seeking a firmware or embedded software engineering role to apply strong DSP fundamentals and a drive to ship reliable, resource-constrained hardware products.

Embedded Systems Intern at BIT Mesra — Birla Institute of Technology (2025-06 – 2025-07)

• Built a standalone, real-time Parkinson's disease speech detection system on a single STM32F401RE Nucleo-64 (Cortex-M4 @ 84 MHz, 96 kB SRAM) with no OS, no cloud, and no external processor.
• Implemented dual bare-metal feature-extraction pipelines in Embedded C: a 6-stage MFCC pipeline (ADC capture, Hamming windowing, 1024-pt FFT via CMSIS-DSP, 26-band Mel filterbank, log compression, DCT) and an STFT + rank-1 NMF pipeline, producing a 29-dimensional feature vector per 224 ms audio window in under 8 ms of DSP compute.
• Integrated a MAX4466 electret microphone on ADC1 and an SSD1306 OLED (I2C at 400 kHz) for fully on-device classification output, eliminating all host-PC dependency.
• Embedded NanoEdge AI 2-class KNN inference library into firmware Flash; verified DSP correctness against a Python reference implementation (max deviation less than 10e-3 across all 29 features).

B.Tech in Electronics & Communication Engineering – Babasaheb Bhimrao Ambedkar University (2022 – 2026)

Intermediate (10+2) in Science – M.G.M. Senior Secondary School (2022-01)

Matriculation (10th) – M.G.M. Senior Secondary School (2020-01)