Architect distributed fine-tuning pipelines for small encoder and decoder models
Implement LoRA, adapters, distillation, and compression workflows
Design experiment tracking, reproducibility, and dataset versioning systems
Optimize training efficiency (GPU utilization, memory, throughput, cost)
Design scalable RL training workflows (policy optimization, reward modeling)
Integrate RL with supervised fine-tuning and distillation
Build evaluation loops and automated regression detection
Build scalable ingestion pipelines for structured and unstructured data
Design dataset curation, filtering, and quality enforcement systems
Implement reproducible data workflows tied to training runs
Architect low-latency inference services
Design safe production deployment workflows

AI Platform Engineer

Key skills