Staff Engineer - ML Platform

Delivery Hero

Responsibilities

• Design, build, and maintain scalable, reusable, and reliable ML platforms and tooling that support the entire ML lifecycle, including data ingestion, model training, evaluation, deployment, and monitoring for both traditional and generative AI models.
  

• Develop standardized ML workflows and templates using MLflow and other platforms, enabling rapid experimentation and deployment cycles.
  

• Implement robust CI/CD pipelines, Docker containerization, model registries, and experiment tracking to support reproducibility, scalability, and governance in ML and genAI.
  

• Collaborate closely with genAI experts to integrate and optimize genAI technologies, including transformers, embeddings, vector databases (e.g., Pinecone, Redis, Weaviate), and real-time retrieval-augmented generation (RAG) systems.
  

• Automate and streamline ML and genAI model training, inference, deployment, and versioning workflows, ensuring consistency, reliability, and adherence to industry best practices.
  

• Ensure reliability, observability, and scalability of production ML and genAI workloads by implementing comprehensive monitoring, alerting, and continuous performance evaluation.
  

• Integrate infrastructure components such as real-time model serving frameworks (e.g., TensorFlow Serving, NVIDIA Triton, Seldon), Kubernetes orchestration, and cloud solutions (AWS/GCP) for robust production environments.
  

• Drive infrastructure optimization for generative AI use-cases, including efficient inference techniques (batching, caching, quantization), fine-tuning, prompt management, and model updates at scale.
  

• Partner with data engineering, product, infrastructure, and genAI teams to align ML platform initiatives with broader company goals, infrastructure strategy, and innovation roadmap.
  

• Contribute actively to internal documentation, onboarding, and training programs, promoting platform adoption and continuous improvement.

Requirements

Technical Experience

• Strong software engineering background with experience in building distributed systems or platforms designed for machine learning and AI workloads.
  

• Expert-level proficiency in Python and familiarity with ML frameworks (TensorFlow, PyTorch), infrastructure tooling (MLflow, Kubeflow, Ray), and popular APIs (Hugging Face, OpenAI, LangChain).
  

• Experience implementing modern MLOps practices, including model lifecycle management, CI/CD, Docker, Kubernetes, model registries, and infrastructure-as-code tools (Terraform, Helm).
  

• Demonstrated experience working with cloud infrastructure, ideally AWS or GCP, including Kubernetes clusters (GKE/EKS), serverless architectures, and managed ML services (e.g., Vertex AI, SageMaker).
  

• Proven experience with generative AI technologies: transformers, embeddings, prompt engineering strategies, fine-tuning vs. prompt-tuning, vector databases, and retrieval-augmented generation (RAG) systems.
  

• Experience designing and maintaining real-time inference pipelines, including integrations with feature stores, streaming data platforms (Kafka, Kinesis), and observability platforms.
  

• Familiarity with SQL and data warehouse modeling; capable of managing complex data queries, joins, aggregations, and transformations.
  

• Solid understanding of ML monitoring, including identifying model drift, decay, latency optimization, cost management, and scaling API-based genAI applications efficiently.