AI Platform & Cloud Engineer

Axle•Rockville, MD

19h

About The Position

The AI Platform & Cloud Engineer will help sustain the hybrid cloud production environment for the SOM Center’s data ecosystem. This role serves as the technical interface between Data Science and IT, focusing on Platform Engineering: building the internal developer platform (IDP) that utilizes the IT-managed Kubernetes infrastructure and cloud resource to scale resources for workflow orchestration, knowledge graph data pipelines, and distributed model inference.

Requirements

Bachelor’s or master’s degree in computer science or engineering with experience in Cloud Engineering, MLOps, or SRE.
Proficiency in Python and Infrastructure as Code concepts, with experience in major cloud platforms (GCP preferred, or AWS).
AI Productivity: Demonstrated ability to leverage AI-driven coding assistants and LLMs to increase development velocity and code quality.
Experience utilizing Hybrid Cloud architectures and configuring workloads for burst computing (Spot instances, Autoscaling groups).
Experience refactoring research-grade code into production-grade services (Docker/Kubernetes).
Experience with Workflow Orchestration tools (Airflow, Prefect, or Dagster) and Vector Database administration.

Nice To Haves

Experience deploying applications to Kubernetes (GKE/EKS) and using GitOps workflows (ArgoCD/Flux).
Knowledge of Graph Database administration (Neo4j) and object storage architectures.
Familiarity with Serverless event processing (Cloud Functions) and ML Engineering concepts (quantization, distillation, serving via Triton/vLLM).

Responsibilities

IT Collaboration & K8s Support: Collaborate closely with the dedicated IT team to define compute requirements and orchestrate workloads on the new Kubernetes cluster. The engineer will not manage the cluster directly but will ensure data science applications are correctly containerized and configured to run efficiently on the infrastructure provided by IT.
Infrastructure Strategy: Define the Infrastructure as Code (IaC) specifications for application-level resources, working with IT to ensure on-premises GPU clusters and public cloud environments (GCP/AWS) are utilized effectively.
Refactoring & Model Serving : Transform experimental code (Jupyter Notebooks, R scripts) developed by NLP and Omics researchers into robust, containerized software packages. Deploy and optimize model inference servers (e.g., vLLM, Triton Inference Server) to expose AI models as reliable internal APIs.
Workflow Orchestration: Deploy and maintain the Workflow Orchestration platform (e.g., Apache Airflow, Prefect, or Dagster) to manage dependencies between data ingestion, model inference, and state updates, serving as the central execution controller for distributed processes.
AI-Assisted Development: Actively utilize AI-assisted coding tools (e.g., GitHub Copilot) to accelerate code generation, documentation, and refactoring processes to increase overall productivity.
Data Foundation: Administer the Data Foundation infrastructure, including supporting Graph Databases (e.g., Neo4j), Vector Databases (e.g., Milvus, pgvector) for RAG implementations, and ETL pipelines to ingest massive public datasets (e.g., Human Cell Atlas) into the Data Lake.
Cloud Agent Architecture: Architect and deploy managed Cloud AI Agents (e.g., via Vertex AI) to orchestrate complex reasoning workflows, including and not limited to parsing scientific literature, querying omics databases, and validating experimental protocols against Knowledge Graphs.
Security Implementation: Collaborate with data scientists to implement Workload Identity federation and secrets management (e.g., Vault), ensuring automated workflows securely authenticate against enterprise resources managed by IT.