Senior LLM-MLops Engineer @ Square One Resources

Senior MLOps/LLMOps Engineer @ Square One Resources – Rzeszów, Poland As a Senior MLOps/LLMOps Engineer , you will be at the forefront of building and scaling our AI/ML infrastructure, bridging the gap between cutting‑edge large language models and production‑ready systems. You will play a pivotal role in designing, deploying, and operating the platforms that power our AI‑driven products, working at the intersection of DevOps, MLOps, and emerging LLM technologies. In this role, you'll architect robust, scalable infrastructure for deploying and monitoring large language models (LLMs) such as GPT and Claude‑family models in AWS Bedrock & AWS AI Foundry, while ensuring security, observability, and reliability across multi‑tenant ML workloads. You will collaborate closely with data scientists, ML engineers, platform teams, and product stakeholders to create seamless, self‑serve experiences that accelerate AI innovation across the organization. Responsibilities Run and evolve our ML/LLM compute infrastructure on Kubernetes/EKS (CPU/GPU) for multi‑tenant workloads, ensuring portability across AWS and Azure AI Foundry regions with region‑aware scheduling, cross‑region data access, and artifact management. Engage with platform and infrastructure teams to provision and maintain access to cloud environments (AWS, Azure), ensuring seamless integration with existing systems. Set up and maintain deployment workflows for LLM‑powered applications, handling environment‑specific configurations across development, staging/UAT, and production. Build and operate GitOps‑native delivery pipelines using GitLab CI, Jenkins, ArgoCD, Helm, and FluxCD to enable fast, safe rollouts and automated rollbacks. Deploy, scale, and optimize large language models (GPT, Claude, and similar) with deep consideration for prompt engineering, latency/performance tradeoffs, and cost efficiency. Operate and maintain Argo Workflows as reliable, self‑serve orchestration platforms for data preparation, model training, evaluation, and large‑scale batch compute. Implement and evaluate models using AI Observability frameworks to track model performance, drift, and quality in production. Design and maintain robust CI/CD pipelines with isolated development, staging, and production environments to support safe iteration, reproducibility, and full lifecycle observability. Implement Infrastructure as Code (IaC) using Terraform, CloudFormation, and Helm to automate provisioning, configuration, and scaling of cloud resources. Manage container orchestration, secrets management (e.g., AWS Secrets Manager), and secure deployment practices across all environments. Set up and analyze comprehensive observability stacks using Prometheus, Grafana, and Splunk to monitor model health, infrastructure performance, and system reliability. Qualifications 8+ years of experience in DevOps, Platform Engineering, or Site Reliability Engineering, with at least 2+ years focused on MLOps/LLMOps. Deep hands‑on expertise with AWS services, including Bedrock, S3, EC2, EKS, RDS/PostgreSQL, ECR, IAM, Lambda, Step Functions, and CloudWatch. Production experience managing Kubernetes workloads in EKS, including GPU workloads, auto‑scaling, resource quotas, and multi‑tenant configurations. Proficient in container orchestration (Docker, Kubernetes), secrets management, and implementing GitOps‑style deployments using Jenkins, ArgoCD, FluxCD, or similar tools. Practical understanding of deploying and scaling LLMs (e.g., GPT and Claude‑family models), including prompt engineering, latency/performance tradeoffs, and model evaluation. Strong programming skills in Python (FastAPI, Django, Pydantic, boto3, Pandas, NumPy) with solid computer science fundamentals. Working knowledge of machine learning techniques and frameworks (e.g., scikit‑learn, TensorFlow, PyTorch). Experience building and operating data pipelines with principles of idempotency, retries, backfills, and reproducibility. Expertise in Infrastructure as Code (IaC) using Terraform, CloudFormation, and Helm. Proven track record designing and maintaining CI/CD pipelines with GitLab CI, Jenkins, or similar tools. Observability experience with Prometheus, Grafana, Splunk, Datadog, Loki/Promtail, OpenTelemetry, and Sentry, including implementing sensible alerting strategies. Strong grasp of networking, security concepts, and Linux systems administration. Excellent communication skills with ability to collaborate across development, QA, operations, and product teams. Self‑motivated, proactive, with a strong sense of ownership and a passion for removing friction and improving developer experience. Nice to Have Experience with distributed compute frameworks such as Dask, Spark, or Ray. Familiarity with NVIDIA Triton, TorchServe, or other inference servers. Experience with ML experiment tracking platforms like Weights & Biases, MLflow, or Kubeflow. FinOps best practices and cost attribution strategies for multi‑tenant ML infrastructure. Exposure to multi‑region and multi‑cloud designs, including dataset replication strategies, compute placement, and latency optimization. Experience with LakeFS, Apache Iceberg, or Delta Lake for data versioning and lakehouse architectures. Knowledge of data transformation tools such as DBT. Experience with data pipeline orchestration tools like Airflow or Prefect. Familiarity with Snowflake or other cloud data warehouses. Understanding of responsible AI practices, model governance, and compliance frameworks. #J-18808-Ljbffr