Embedded Edge AI / ML Engineer

Tosil System; Thiruvananthapuram; On-site; INR 2,500,000 - 4,000,000; Full time

Benefits offered by this job

Silicon‑proven projects

Custom AI accelerators

End‑to‑end AI system design

Collaborative environment

Job summary

TOSIL Systems seeks an experienced Embedded Edge AI/ML Engineer to deploy and optimize DL workloads on embedded and accelerator platforms, spanning audio AI, ASR, and low‑power GenAI inference on custom SoCs.

You will optimize models, work across chip design teams, and leverage GPUs, NPUs, DSPs, FPGAs, and RISC‑V platforms to deliver latency‑sensitive AI at the edge. Kerala‑based opportunity, strong hardware/software co‑design emphasis.

Qualifications

• 5+ years of Embedded Systems and Edge AI/ML development.
• Proficiency in C/C++ and Python.
• Hands-on with TensorFlow, TensorFlow Lite, Keras, PyTorch, ONNX.

Responsibilities

• Design, optimize, and deploy Deep Learning and Gen‑AI inference pipelines for edge and embedded systems.
• Work on audio AI workloads such as keyword spotting, ASR, speech enhancement, and audio classification.
• Deploy and optimize ML/DL models using TensorFlow, TensorFlow Lite, Keras, PyTorch, and ONNX.
• Perform model optimization using quantization (INT8/INT4), pruning, sparsity, and mixed‑precision techniques.
• Enable and deploy TinyML models on ultra‑low‑power MCUs and DSP‑class processors.
• Integrate and optimize AI workloads on custom Edge AI accelerators, NPUs, DSPs, GPUs, and FPGA accelerators.
• Enable AI workloads on RISC‑V based SoCs and heterogeneous architectures.
• Work with NVIDIA platforms (Jetson, CUDA, TensorRT, cuDNN) for edge and HPC‑class AI inference.
• Collaborate with silicon architecture, RTL, FPGA, and verification teams to co‑design AI hardware and software.
• Profile, benchmark, and optimize latency, throughput, power, and memory for silicon‑proven designs.
• Develop supporting middleware, runtime software, and drivers in C/C++ and Python.
• Validate AI accuracy, performance, and robustness on real silicon and FPGA prototypes.
• Contribute to tape‑out‑ready AI software stacks and post‑silicon bring‑up activities.
• Document AI architecture, deployment flows, and optimization methodologies.

Skills

Embedded systems

Python

C/C++

TensorFlow

PyTorch

ONNX

Edge AI

Quantization

NVIDIA Jetson

RISC‑V

Education

B.Tech / B.E. in Electronics/CS/EE

M.Tech / MS in AI / ML (strong plus)

Tools

TensorFlow

TensorFlow Lite

Keras

PyTorch

ONNX

CUDA

Jetson

Yocto

Linux BSPs

Job description

Job Summary

TOSIL Systems is seeking a highly experienced Embedded Edge AI / ML Engineer with 5+ years of hands‑on experience in deploying and optimizing Deep Learning and Generative AI workloads on embedded, edge, and accelerator‑based platforms. This role is closely aligned with TOSIL's silicon‑proven work in audio AI, keyword spotting, speech/ASR, healthcare signal processing, and low‑power Gen‑AI inference on custom AI accelerators and SoCs. The ideal candidate will work across the full stack, from model optimization to deployment on custom silicon, Edge AI accelerators, RISC‑V platforms, and HPC‑class systems.

Responsibilities

• Design, optimize, and deploy Deep Learning and Gen‑AI inference pipelines for edge and embedded systems.
• Work extensively on audio AI workloads such as keyword spotting, ASR front‑ends, speech enhancement, and audio classification.
• Deploy and optimize ML/DL models using TensorFlow, TensorFlow Lite, Keras, PyTorch, and ONNX.
• Perform model optimization using quantization (INT8/INT4), pruning, sparsity, and mixed‑precision techniques.
• Enable and deploy TinyML models on ultra‑low‑power MCUs and DSP‑class processors.
• Integrate and optimize AI workloads on custom Edge AI accelerators, NPUs, DSPs, GPUs, and FPGA‑based accelerators.
• Enable AI software stacks on RISC‑V based SoCs and heterogeneous architectures.
• Work with NVIDIA platforms (Jetson, CUDA, TensorRT, cuDNN) for edge and HPC‑class AI inference.
• Collaborate closely with silicon architecture, RTL, FPGA, and verification teams to co‑design AI hardware and software.
• Profile, benchmark, and optimize latency, throughput, power, and memory for silicon‑proven designs.
• Develop supporting middleware, runtime software, and drivers in C/C++ and Python.
• Validate AI accuracy, performance, and robustness on real silicon and FPGA prototypes.
• Contribute to tape‑out‑ready AI software stacks and post‑silicon bring‑up activities.
• Document AI architecture, deployment flows, and optimization methodologies.

Requirements

• 5+ years of experience in Embedded Systems and Edge AI / ML development.
• Strong proficiency in C/C++ and Python.
• Hands‑on experience with Deep Learning frameworks: TensorFlow, TensorFlow Lite, Keras, PyTorch, ONNX.
• Strong understanding of Deep Learning architectures including CNNs, RNNs, Transformers, and audio‑specific models.
• Proven experience with TinyML workflows and ultra‑low‑power AI deployment.
• Experience deploying AI models on Embedded Linux and/or RTOS platforms.
• Strong experience with custom AI accelerators, NPUs, DSPs, GPUs, or FPGA‑based accelerators.
• Hands‑on experience with NVIDIA AI stack: Jetson, CUDA, TensorRT.
• Experience enabling AI workloads on RISC‑V based platforms.
• Solid understanding of hardware/software co‑design and AI accelerator interfaces.
• Experience with profiling, benchmarking, and debugging AI systems on real silicon.
• Familiarity with HPC concepts and high‑throughput inference systems.
• Proficiency with Git and collaborative development workflows.

Preferred qualifications

• Experience with Gen‑AI inference optimization (LLMs, diffusion models, audio transformers).
• Prior involvement in silicon‑proven designs or post‑silicon bring‑up.
• Experience with Yocto, Linux BSPs, and system‑level optimization.
• Knowledge of power‑aware, thermal‑aware, and low‑latency AI system design.
• Publications, patents, or open‑source contributions in AI / ML or Edge AI.
• Strong analytical and cross‑functional collaboration skills.

Benefits

• Opportunity to work on silicon‑proven Embedded AI and Gen‑AI systems.
• Direct involvement in custom AI accelerator and SoC development.
• Exposure to end‑to‑end AI system design from model to silicon.
• Collaborative environment with experts in AI, hardware architecture, and RTL design.

Qualifications

• B.Tech / B.E. in Electronics, Computer Science, Electrical Engineering, or a related discipline.
• M.Tech / MS with specialization in AI / ML is a strong plus.

Disclaimer: This job posting has been aggregated from external source. Role details, content, and availability are subject to change. Applicants are advised to confirm the latest information directly on the company website before applying.