MLOps / AI Ops Engineer

We are seeking an experienced MLOps / AI Ops Engineer to join our newly established DevOps & Run Team in the Data & AI CoE at Prague's NN Digital Hub. Shape the future of enterprise AI/LLMOps , collaborate across borders, drive innovation, and grow with a competitive package and continuous learning.

Your impact at NN

As MLOps / AIOps Engineer, you'll design pipelines, operationalize ML/LLM solutions, and ensure they run securely and cost-efficiently. You'll work with Data Scientists and DevOps Engineers to convert experiments into production-ready pipelines, strengthen observability, and drive responsible AI adoption. You will operationalize AI/ML and LLM solutions on our Cloud Lakehouse (Databricks on top of Azure), build pipelines, productionize models, and ensure scalability, security, and cost efficiency. Experience with next-gen frameworks (e.g. Mosaic AI, LangChain) is a plus, and you'll have opportunities to learn them on the job.

The tasks you will perform

Pipelines & Deployment

• Design and implement CI/CD pipelines for ML and LLM workloads using Azure DevOps, Databricks Repos, and Asset Bundles.
• Automate model packaging, registration, versioning, and promotion with MLflow and Unity Catalog.
• Manage configurations (YAML/JSON) for orchestration, maintain Git-based version control, and enforce automation best practices.
• Support scalable feature engineering and feature store integration for reusability across teams.

Model Operations

• Operationalize pre-trained and vendor LLMs/SLMs (e.g., Llama, Mistral, Azure OpenAI, Anthropic) in production environments.
• Enable both batch scoring and real-time inference with Databricks, AKS, or vendor APIs.
• Build and maintain Retrieval-Augmented Generation (RAG) pipelines with vector search solutions (e.g. Databricks Vector Search, Azure AI Search, FAISS).
• Contribute to Python library development and packaging (e.g., Poetry, wheels) for internal reuse.

Monitoring, Reliability & Responsible AI

• Establish observability for data drift, concept drift, pipeline health, and model performance (accuracy, latency, throughput), including LLM-specific metrics like hallucination rates and token efficiency.
• Set up monitoring and alerting for agents, endpoints, and workflows using tools like Azure Monitor, Prometheus, Grafana, and Databricks dashboards — with automated remediation for failures.
• Define and maintain SLOs/SLIs for AI services (e.g., uptime, retraining cadence), and ensure reliability across batch and real-time systems.
• Implement explainability (e.g., SHAP, MLflow), enable human-in-the-loop reviews, and support compliance with GDPR, the EU AI Act, and ethical AI principles through policy-as-code and audit dashboards.

Platform Operations & Continuous Improvement

• Configure and optimize Databricks components (clusters, DLT, Unity Catalog, Feature Store) for performance and cost efficiency.
• Deploy hybrid inference setups using Mosaic AI, AKS, and external vendor-hosted LLMs, applying FinOps to manage GPU/CPU usage and cloud spend.
• Build reusable infrastructure modules with Terraform and Bicep, and align with DevOps on shared CI/CD and observability practices.
• Mentor AI squads on advanced MLOps/LLMOps and continuously adopt emerging frameworks

What we expect from you

• Bachelor's/Master's in Computer Science, Data/AI Engineering, or related field.
• 4–8 years in MLOps / Data Engineering / AI Ops
• Hands-on experience operationalizing pre-trained LLMs/SLMs in production.
• Advanced English proficiency required for working in an international environment.

More about technical skills needed:

Must-have:

• Python (incl. packaging), SQL, PySpark
• CI/CD with Azure DevOps or GitHub Actions
• MLflow & model lifecycle management
• Experience deploying ML/LLM models into production

Nice-to-have / Learn on the job:

• Databricks Lakehouse (Delta Lake, Unity Catalog, Mosaic AI)
• LangChain/LangGraph, RAG pipelines, Vector Search (FAISS, Azure AI Search)
• Kubernetes/AKS, Terraform/Bicep

Perks of joining NN

• We allow you to work where you feel the most comfortable, whether it is in the office or from home, and we contribute to your home office expenses every month.
• We understand the importance of having a work-life balance, which is why we offer 5 weeks of vacation, 5 well-being days, additional paid time off for personal and family events, and 1 volunteering day to support our community.
• In addition to your base salary you will have a lump-sum meal allowance, up to CZK 20,000 in the Cafeteria per year, the possibility of arranging a MultiSport card, the possibility of contributing to supplementary pension insurance / supplementary pension savings, and a discount on life insurance.
• We believe that your professional and personal growth is crucial, which is why we provide you with tailor-made professional training.
• Your friends and acquaintances are a valuable source of talent for us, which is why we offer up to 60,000 CZK as a reward for recommending a suitable candidate.
• A business laptop and an iPhone with a paid O2 tariff and a data package are basic tools for your work.

What is hiring process with us like

Once we review your application and English CV, Veronika (Talent Acquisition Partner) will give you a short introductory call to discuss your previous experience and the open position.

If the initial conversation goes well, you'll be invited to two interview rounds. The first round will ideally take place in person at our Prague office (or online) with Karel (Hiring Manager, AI & Automation Manager) and Veronika. In the second round, you'll meet several members of the team you may soon be joining.