Nitesh Singh

I'm a hands-on Embedded AI/ML Engineer with 12+ years of experience building real-time intelligence for safety-critical automotive systems — from Autonomous Driving perception pipelines to ADAS sensor fusion and in-vehicle AI inference on edge hardware. My core expertise sits at the intersection of Machine Learning, Computer Vision, and Embedded Systems — designing and deploying C++ inference pipelines on QNX/Linux targets, optimizing models for NPU/GPU execution using Qualcomm AI Runtime (QNN), and integrating Deep Learning into constrained, latency-sensitive environments. What I've shipped: .

ML perception modules for BMW iX and Tata Safari/Harrier ADAS systems (Aptiv). Real-time AI inference on automotive-grade NPU/GPU hardware (BMW TechWorks). Radar + Camera sensor fusion with Kalman filtering for object detection and tracking.

Audio streaming and embedded validation systems on QNX/Linux (Bosch).

Computer

Vision pipelines for Autonomous Driving across 5+ years at KPIT (India & Germany) I thrive in IC roles where I can go deep — architecting systems, optimizing models, and solving hard embedded AI problems close to the metal.

Tech: C++ · Python · QNX · Linux · Qualcomm QNN · Computer Vision · Deep Learning · ADAS · CAN · Gtest · Kalman Filtering · Radar

12+ years of hands-on experience in automotive software engineering, progressively specializing in AI/ML, Embedded Systems, Autonomous Driving, and ADAS.

Career arc: Started as a C++ engineer in Pune,India (2013), grew through R&D and embedded software roles, relocated to Munich, Germany with KPIT, and have since worked with Bosch, Aptiv, and BMW TechWorks — three of the most respected names in automotive engineering.

Domain expertise:

• Embedded AI/ML — deploying inference pipelines on NPU/GPU using Qualcomm QNN
• ADAS & Autonomous Driving — sensor fusion, Radar + Camera perception, Kalman filtering, object detection
• Computer Vision & Deep Learning — production ML models on real vehicles (BMW iX, Tata Safari/Harrier)
• Performance Engineering — low-latency, high-reliability C++ on QNX/Linux targets

Companies: BMW TechWorks → Aptiv → Bosch → KPIT (India + Germany) → TouchMagix → SmartLeaven → Pingaala

Core stack: C++ · Python · QNX · Linux · Qualcomm QNN · NPU · GPU · CAN · Gtest · Deep Learning · Computer Vision

What makes me stand out: I've shipped real AI on real vehicles — not just research or prototypes. That's rare and highly valuable for Embedded AI/ML IC roles.

Mtech in AI/ML (Nov, 2024 - Present) from BITS Pilani, India

B.E in Computer Science Engineering ( 2009-2013) from KLESCET,India