23 Aug
|
Sandisk
|
Bengaluru
Job Description
We are seeking an experienced AI Software Stack Test Lead to define, drive, and execute validation strategy for a next-generation AI computational storage platform. This role owns end-to-end verification of the AI software stack, including neural network operators, compiler-generated workloads, runtime execution, memory management, scheduling systems, and model-level performance.
The ideal candidate combines solid software validation expertise with deep knowledge of AI/ML execution pipelines and will lead validation efforts across compiler, runtime, kernel, and hardware teams to ensure correctness, scalability, reliability, and performance.
Key Responsibilities
Validation Strategy & Technical Leadership
Own and define the end-to-end validation strategy for the AI software stack.
Establish validation methodologies, coverage metrics, quality gates, and release-readiness criteria.
Lead validation efforts across operator libraries, runtime systems, execution frameworks, and AI workloads.
Mentor engineers and drive best practices for automation, performance validation, and debugging.
Functional Validation
Validate neural network operators and compute kernels
Validate graph-level execution and end-to-end model inference workflows.
Verify numerical correctness against reference frameworks such as PyTorch or TensorFlow.
Ensure behavioral consistency across software releases and hardware revisions.
Runtime & Execution Validation
Validate memory allocation and movement across host and device settings.
Execute stress, concurrency, and multi-device validation scenarios.
Performance & Scalability Validation
Design and maintain benchmarking frameworks for:
Operator-level performance
Model-level performance
Latency and throughput
Time-to-first-token/result
Multi-device scaling efficiency
📌 Staff Test Engineer Ai Sw Stack 9 13 Years Bengaluru
🏢 Sandisk
📍 Bengaluru