Full Stack Machine Learning Engineer
Salary not listed • Bengaluru, IND • Full-time
Posted 22d ago
Remote Work Policy
On-site
Employment Type
Full-time
Categories
Machine Learning Engineer
About the job
We are seeking an ML Engineer to transition AI models from reference implementations to efficient execution on specialized AI accelerator hardware. This role involves owning the model bring-up process, developing MLIR-based lowering paths, ensuring numerical correctness, and optimizing performance across the entire stack, including the model, compiler, kernels, and runtime. The position requires a blend of practical knowledge of model architectures and strong compiler and systems engineering expertise.
Responsibilities
- Bring up new models by understanding architectures, loading weights, implementing execution paths, and verifying correctness.
- Develop and extend MLIR dialects, graph transformations, lowering passes, and hardware-specific mappings to lower models to hardware.
- Enable key model operations like attention, matrix multiplication, and normalization through compiler and kernel modifications.
- Optimize execution by improving operator fusion, tensor layouts, tiling, memory allocation, data movement, and parallel execution.
- Tune prefill and decode performance, KV cache management, batching, and quantization for improved inference latency, throughput, and memory efficiency.
- Validate model quality by investigating numerical differences and assessing the accuracy impact of precision changes and optimizations.
- Diagnose bottlenecks using profiling, execution traces, and hardware counters to identify compute, memory, communication, and runtime limitations.
- Collaborate with hardware, compiler, kernel, and runtime teams to ensure reliable model support and repeatable performance benchmarks.
Requirements
- Strong programming skills in C++ and Python.
- Hands-on experience bringing up and debugging ML models in PyTorch or a similar framework.
- Practical experience with MLIR, including dialects, rewrite patterns, transformation passes, and lowering pipelines.
- Understanding of compiler fundamentals such as intermediate representations, dataflow analysis, and code generation.
- Knowledge of transformer architectures, attention mechanisms, tensor operations, and numerical precision.
- Experience profiling and optimizing workloads on GPUs or other AI accelerators.
- Ability to debug correctness and performance issues across model code, compiler-generated code, kernels, and runtime execution.
Benefits
- Build a breakthrough AI platform beyond the constraints of the GPU.
- Publish and open source their cutting-edge AI research.
- Work on one of the fastest AI supercomputers in the world.
- Enjoy job stability with startup vitality.
- A simple, non-corporate work culture that respects individual beliefs.