Engineer, Principal/manager - Machine Learning, Ai Job in Qualcomm India Private Limited

Engineer, Principal/Manager - Machine Learning, AI

Location: Bangalore, Karnataka, India
Company: Qualcomm India Private Limited

General Summary

Qualcomm is seeking an experienced and visionary Principal AI/ML Engineer to lead research, development, and optimization of AI inference systems. This role involves developing high-performance AI models, optimizing deployments across various hardware platforms, and contributing to research in model compression, quantization, and hardware-aware optimization.

Education & Experience

• PhD with 6+ years, Master's with 7+ years, or Bachelor's with 8+ years in Engineering, CS, or related field.
• 20+ years of experience in AI/ML development; 5+ years in inference optimization and debugging.

Key Responsibilities

Model Optimization & Quantization

• Optimize models using quantization (INT8, INT4, mixed precision), pruning, and knowledge distillation.
• Implement PTQ and QAT techniques for deployment.
• Experience with TensorRT, ONNX Runtime, OpenVINO, TVM.

AI Hardware Acceleration & Deployment

• Target platforms: Hexagon DSP, CUDA GPUs, TPUs, NPUs, FPGAs, Habana Gaudi, Apple Neural Engine.
• Use Python APIs: cuDNN, XLA, MLIR for hardware acceleration.
• Benchmark and debug performance across platforms.

AI Research & Innovation

• Research on efficient AI inference: model compression, low-bit precision, sparse computing.
• Explore architectures like Sparse Transformers, Mixture of Experts, Flash Attention.
• Publish in ML conferences: NeurIPS, ICML, CVPR; contribute to open-source projects.

Technical Expertise

• Optimization of LLMs, LMMs, LVMs for inference.
• Deep Learning frameworks: TensorFlow, PyTorch, JAX, ONNX.
• Expert in CUDA, cuPy, Numba, TensorRT, ONNX Runtime, OpenVINO.
• Skilled in Python for scalable AI development.
• Experience with ML runtime delegates: TFLite, ONNX, Qualcomm AI Stack.
• Debugging: Netron, TensorBoard, PyTorch Profiler, Nsight, perf, Py-Spy.
• Cloud inference: AWS Inferentia, Azure ML, GCP AI Platform, Habana Gaudi.
• Hardware-aware optimization: oneDNN, ROCm, MLIR, SparseML.
• Contributions to open-source and research publications are a strong plus.

Leadership & Collaboration

• Lead a team of engineers in Python-based AI inference and optimization.
• Collaborate with researchers, software engineers, DevOps, and hardware vendors.
• Define debugging, deployment, and performance tuning best practices.