ML Compiler & SDK Validation Engineer

Brief About Tech Mahindra – Silicon Engineering
Tech Mahindra Silicon Engineering is a key part of Tech Mahindra’s semiconductor and engineering services portfolio, delivering end-to-end solutions across the silicon lifecycle for global semiconductor, automotive, consumer, networking, and industrial clients. The practice focuses on VLSI, Embedded Systems, SoC Design, AI/ML Enablement, and Platform Engineering, supporting customers from architecture and RTL design through implementation, verification, validation, software enablement, and post-silicon support.
The Silicon Engineering unit works across IP, Sub-System, SoC, and AI Accelerator Development, offering deep expertise in areas such as Design Verification, Physical Design, Embedded Software, AI/ML Compiler Validation, SDK Enablement, Performance Optimization, and System Engineering. Teams leverage industry-leading EDA tools, AI frameworks, compiler technologies, and hardware acceleration platforms to deliver high-performance and scalable solutions.
Backed by Tech Mahindra’s global engineering DNA, the Silicon Engineering practice collaborates with leading semiconductor and technology companies worldwide, serving as an extended engineering arm to accelerate silicon and software development programs, optimize product quality, and enable faster time-to-market for next-generation AI-enabled systems.
About the Role – ML Compiler & SDK Validation Engineer
In this exciting role, you will be responsible for validating ML Compiler toolchains, SDK software stacks, AI frameworks, and machine learning model execution flows across advanced AI acceleration platforms. You will work closely with Compiler, SDK, AI Framework, Software, Silicon Validation, and Architecture teams to ensure functional correctness, performance, model accuracy, and software readiness.
The role involves ML model evaluation, compiler validation, SDK testing, performance benchmarking, regression automation, debugging, and validation signoff activities across AI/ML workloads deployed on CPU, GPU, DSP, NPU, and custom AI accelerators.
Responsibilities – Key Responsibilities
Validate ML Compiler toolchains and SDK software stacks across various AI/ML workloads.
Develop and execute comprehensive validation plans, test strategies, test cases, and regression suites for compiler and SDK releases.
Perform ML model validation and accuracy evaluation across TensorFlow, PyTorch, ONNX, and other AI frameworks.
Validate model compilation flows including graph conversion, optimization, quantization, and deployment pipelines.
Execute functional, accuracy, performance, stress, and regression testing of compiler-generated outputs.
Analyze and debug compiler failures, model accuracy mismatches, runtime errors, performance issues, and SDK integration problems.
Perform benchmarking of AI workloads and analyze metrics such as latency, throughput, memory utilization, and inference efficiency.
Work closely with compiler development, architecture, framework, and silicon teams to root-cause and resolve issues.
Validate quantization flows including INT8, FP16, BF16, and mixed-precision deployments.
Develop automation frameworks for validation, regression execution, result comparison, and reporting.
Support CI/CD integration for compiler and SDK validation environments.
Generate validation reports, coverage metrics, issue tracking documentation, and release signoff reports.
Participate in design reviews, technical discussions, customer interactions, and methodology improvements.
Collaborate with global engineering teams to ensure high-quality compiler and SDK releases aligned with project milestones.
Qualifications
BE/BTech/ME/MTech degree in Computer Science, Electronics, Electrical Engineering, Artificial Intelligence, Machine Learning, Computer Engineering, Data Science, or related engineering disciplines.
Required Skills
Strong experience in ML Compiler Validation, SDK Validation, AI Software Validation, or Performance Validation.
Hands-on expertise in validating AI/ML models using TensorFlow, PyTorch, ONNX, Keras, or similar frameworks.
Good understanding of ML Compiler architectures, graph optimizations, inference engines, and deployment workflows.
Experience in model accuracy validation, performance benchmarking, and regression testing.
Knowledge of compiler technologies such as TVM, TensorRT, OpenVINO, MLIR, XLA, Glow, or equivalent compiler frameworks.
Experience validating ML workloads on GPU, DSP, NPU, AI Accelerators, or Heterogeneous Compute Platforms.
Strong debugging skills involving accuracy discrepancies, numerical precision issues, compiler crashes, and SDK integration failures.
Knowledge of quantization techniques including INT8, FP16, BF16, and Mixed Precision.
Proficiency in Python, C++, Shell Scripting, and Automation Framework Development.
Experience with CI/CD validation environments and automated regression execution.
Strong analytical, debugging, problem-solving, communication, and stakeholder management skills.
Ability to work effectively with cross-functional global engineering teams in fast-paced AI and semiconductor development programs.
Preferred Experience
5 to 10 years of experience in ML Compiler Validation, SDK Validation, AI Software Validation, or AI Platform Enablement.
Experience with AI workloads such as Computer Vision, Generative AI, LLMs, Recommendation Systems, Speech Processing, and Deep Learning models.
Exposure to compiler optimization validation, inference engine validation, and accelerator software stack testing.
Experience in AI/ML performance benchmarking and profiling using industry-standard tools.
Familiarity with containerized development environments, CI/CD pipelines, and cloud-based AI validation workflows.
Location
Bangalore
Experience Range
5–10 Years

Work arrangement
No