We are seeking an experienced AI Validation Engineer to lead validation, benchmarking, and quality assurance activities for AI/ML software stacks running on embedded and heterogeneous computing platforms. The ideal candidate will possess strong expertise in AI frameworks, ROCm ecosystems, Linux-based environments, performance analysis, and automation. This role will drive end-to-end AI pipeline validation while collaborating closely with architecture, compiler, runtime, driver, and hardware teams to ensure production-quality AI solutions.

Key Responsibilities :

AI/ML Validation & Quality Ownership :

– Lead validation efforts for complex AI/ML compute stacks across multiple hardware and software platforms.

– Define validation strategies, test plans, methodologies, and quality metrics for AI software and system pipelines.

– Own the complete defect lifecycle, including issue reporting, triage, root-cause analysis, tracking, and closure.

– Ensure comprehensive coverage across functional, performance, regression, stress, scalability, and reliability testing.

End-to-End AI Pipeline Validation :

– Validate complete AI workflows across training, optimization, and inference pipelines.

– Validate ROCm libraries and AI software stack functionality.

– Verify :

1. Model training, conversion, and optimization workflows (e.g., PyTorch to ONNX)

2. Inference runtimes such as ONNX Runtime, TensorRT, ROCm/HIP, and OpenVINO

3. AI compilers and toolchains including TVM, Vitis AI, XDNA, and XLA

4. Kernel execution, memory movement, inference correctness, and accuracy

– Validate AI workload stability, performance, and correctness on Ubuntu and Yocto-based Linux platforms.

AI Benchmarking, Profiling & Performance Optimization :

– Define and execute benchmarking strategies for AI training and inference workloads.

– Profile AI models to identify compute, memory, throughput, and latency bottlenecks.

– Collaborate with compiler, runtime, and hardware teams to drive system-level and model-level optimizations.

– Validate performance improvements across :

1. Model architectures

2. Batch sizes

3. Precision modes (FP32, FP16, INT8)

4. Execution paths and hardware configurations

– Ensure performance regressions are detected early and release performance targets are consistently achieved.

AI Framework & Compute Stack Validation :

– Validate functionality, integration, and performance of AI frameworks including :

1. PyTorch

2. TensorFlow

3. ONNX Runtime

– Execute and validate workloads across heterogeneous compute environments utilizing :

1. ROCm/HIP

2. CUDA

3. OpenCL

4. AI accelerators

– Analyze the impact of framework, compiler, and runtime changes on real-world AI workloads.

Automation & Tool Development :

– Design and develop Python-based validation, benchmarking, and profiling frameworks.

– Build reusable automation for :

1. Test execution

2. Benchmarking

3. Performance profiling

4. Result analysis

5. Reporting and dashboards

– Continuously improve validation efficiency, scalability, and coverage through automation.

Technical Leadership :

– Provide technical leadership and mentorship to validation engineers and junior team members.

– Partner with architecture, compiler, runtime, driver, and hardware teams to resolve functional and performance issues.

– Collaborate effectively with globally distributed cross-functional teams.

– Present validation status, benchmarking results, quality metrics, and performance risks to stakeholders.

Required Skills & Qualifications :

Technical Expertise :

– 8-12 years of experience in AI/ML validation, performance analysis, or software quality engineering.

– Strong understanding of :

1. Deep Learning

2. Large Language Models (LLMs)

3. Recommender Systems

– Strong hands-on experience with ROCm technologies and ROCm stack validation.

– Experience validating AI/ML compute stacks including :

1. HIP

2. CUDA

3. OpenCL

4. OpenVINO

5. PyTorch and TensorFlow ecosystems

– Expertise in end-to-end AI pipeline validation including :

1. Model conversion

2. Inference runtimes

3. AI compilers

4. Kernel execution

5. Accuracy validation

– Advanced Python

– Strong experience in AI benchmarking, profiling, and performance optimization.

– Deep understanding of Linux environments, particularly Ubuntu and Yocto.

Are you interested in this position?

Apply by clicking on the “Apply Now” button below!

#AlbionarcJobs#FintechJobs

#AsiaJobs#MiddleEastCareers

#TechTalent#FintechRecruitment

#FinanceOpportunities#