Work closely with technical marketing, product, and customer support teams in understanding the benchmark data requirements. Explore various bench marking methods , tools used in vision,LLM/VLM models and adopt the same to evaluate Kinara products
Writing and optimizing scripts for deep learning model deployment and benchmarking. Preparing detailed benchmarking and analysis reports for engineering and marketing stakeholders.
Job Responsibility:
Proficiency in Python (for AI workload scripting/optimization), C or C++ (for low-level performance work).
Familiarity with AI accelerators, LLM frameworks, and inference servers.
Measuring and reporting on latency, throughput, accuracy, and power consumption of AI workloads.
Familiarity with MLPerf and MLFlow bench marking suites for computer vision models
Familiarity with various LLM bench mark methods like MMLU, HumanEval, GSM8K, ARC Challenge, GPQA..etc
Working knowledge of Linux environments and device driver internals is often preferred.
Familiarity with CI/CD tools, version control (GitHub), and automation frameworks.
Job Qualification:
BTech in ECE/CS 4+ years of experience, Machine Learning, or a related field.
Exposure to Neural Networks, AI/ML/DL models, training, and frameworks like TensorFlow, Caffe, and PyTorch.
Experience in developing AI workload flows in C++ and Python
Experience in performance evaluation of hardware and software systems.