MLOps Lead Engineer (Dataiku and AWS SageMaker)

Job Title: MLOps Engineer (Dataiku and AWS SageMaker) Location: Reading, Pennsylvania (Onsite – 5 Days/Week at Client Location) Employment Type: Contract / Long-Term Role Overview We are seeking a hands-on MLOps Engineer with strong experience in Dataiku and AWS SageMaker to design, deploy, and operate scalable machine learning and generative AI solutions. The ideal candidate will have experience building agentic AI systems, RAG pipelines, and production-grade ML infrastructure on AWS while ensuring reliability, governance, and performance at scale. This role requires deep expertise in LLMOps, CI/CD automation, containerization, cloud infrastructure, and observability frameworks to support enterprise AI workloads. Key Responsibilities Agentic AI System Design • Design and implement multi-agent architectures including planner, researcher, retriever, executor, and reviewer agents. • Define agent collaboration policies, memory strategies (short/long-term), and tool orchestration frameworks. • Implement supervisor policies and guardrails to ensure safe agent collaboration. Retrieval-Augmented Generation (RAG) Development • Build high-quality RAG pipelines including ingestion, chunking, embeddings, indexing, and retrieval workflows. • Implement evaluation frameworks for precision, recall, groundedness, and hallucination detection. • Ensure proper citation mechanisms and guardrails for enterprise-grade AI applications. AWS-Based AI/ML Production Deployment • Deploy and manage AI solutions using AWS services including: • Amazon Bedrock (Agents, Knowledge Bases, Flows) • AWS Lambda • API Gateway • S3 • DynamoDB • OpenSearch / Vector Databases • Step Functions • CloudWatch • Enable scalable, secure, and fault-tolerant AI systems in production environments. MLOps / LLMOps Implementation • Build automated CI/CD pipelines using GitOps practices. • Implement containerization using Docker and Kubernetes. • Manage Infrastructure as Code (IaC) and deployment pipelines. • Implement secure secrets management, IAM policies, blue-green deployments, and rollback mechanisms. Observability and Model Evaluation • Instrument telemetry including traces, token usage, cost tracking, and latency monitoring. • Build dashboards using Grafana or CloudWatch for operational visibility. • Implement human-in-the-loop review systems, A/B testing, and continuous evaluation pipelines. Reliability and Scalability • Implement caching strategies, queue management, rate limiting, and retry/backoff mechanisms. • Ensure system reliability through idempotency patterns and drift detection mechanisms. • Monitor and optimize system performance under scale. Collaboration and Communication • Work closely with DevOps, Data Engineering, Infrastructure, and Architecture teams. • Document system architectures, SLIs/SLOs, and operational runbooks. • Communicate technical updates and insights to both technical and non-technical stakeholders. Required Qualifications • Bachelor’s degree in Computer Science, Data Science, Engineering, or related field (or equivalent experience). • Proven experience building production-grade MLOps pipelines and AI systems. • Hands-on experience with Dataiku and AWS SageMaker. • Experience designing and deploying RAG pipelines and agent-based AI architectures. • Strong expertise in cloud platforms for AI/ML workloads (AWS preferred). • Solid experience with CI/CD pipelines, Git, Docker, and Kubernetes. • Understanding of model governance, data governance, and AI lifecycle management. • Excellent communication, problem-solving, and collaboration skills. Preferred / Nice to Have Skills • Experience with AWS Bedrock (Agents, Knowledge Bases, Flows). • Experience with OpenSearch or other vector databases. • Familiarity with LangGraph, CrewAI, Semantic Kernel, or AutoGen frameworks. • Experience with Step Functions, Lambda, API Gateway, DynamoDB, and S3. • Knowledge of evaluation frameworks for LLMs including groundedness and hallucination detection. • Dataiku platform expertise including governance, approvals, artifacts, and MLOps deployment flows. Certifications (Nice to Have) • Dataiku ML Practitioner • Dataiku Advanced Designer • Dataiku MLOps Practitioner