Relo: Offered
Required Skills & Experience
- 3–5 years in ML/AI engineering roles owning training and/or serving in production at scale.
- Demonstrated success delivering high-throughput, low-latency ML services with reliability and cost improvements.
- Experience collaborating across Research, Platform/Infra, Data, and Product functions.
- Bachelors in computer science, Electrical/Computer Engineering, or a related field required; Master’s preferred (or equivalent industry experience).
- Strong systems/ML engineering with exposure to distributed training and inference optimization.
Job Description
Insight Global is seeking multipole experienced, driven Machine Learning Engineer to join an established health technology company to join the team in San Jose, CA. This is a full-time, permanent role with competitive salary, bonus, comprehensive benefits, and Relocation.
In this role you'll need:
- Deep Learning Frameworks: Hands-on experience with PyTorch (main focus) and familiarity with TensorFlow.
- Large-Scale Model Training: Exposure to advanced training techniques like Distributed Data Parallel (DDP), Fully Sharded Data Parallel (FSDP), ZeRO, and model parallelism (pipeline/tensor).
- Experience with distributed training is a strong plus. Model Optimization: Skilled in improving model performance through techniques like quantization (PTQ, QAT, AWQ, GPTQ), pruning, knowledge distillation, KV-cache tuning, and using efficient attention mechanisms like Flash Attention.
- Scalable Model Serving: Understanding of how to deploy models at scale, including autoscaling, load balancing, streaming, batching, and caching.
- Comfortable working alongside platform engineers to build robust serving pipelines.
- Data & Storage Systems: Proficient with both SQL and NoSQL databases, vector databases (e.g., FAISS, Milvus, Pinecone, pgvector), and data formats like Parquet and Delta. Familiar with object storage systems.
- Code Quality: Writes efficient, clean, and maintainable code with a focus on performance.
- End-to-End ML Lifecycle: Solid grasp of the full machine learning workflow—from data collection and model training to deployment, inference, optimization, and evaluation.
