Required skills
- Databricks
- Python
- Azure OpenAI
- Retrieval-Augmented Generation (RAG)
- Databricks Mosaic AI
- MLflow
- Mosaic AI Model Serving
- Vector Search
- Databricks Workflows
- Terraform
- GitLab CI/CD
- Unity Catalog
Languages: German
Job description
Develop and operate productive AI agent solutions on Databricks and Azure, from prototype to scalable deployment. Integrate Large Language Models, enterprise data, APIs, and internal systems using RAG, Tool Calling, and agent orchestration.
Project overview
- Conception, development, and operation of productive AI agent solutions on Databricks and
- Azure – from the initial prototype to scalable deployment.
- Development of AI agents and multi-agent workflows for the targeted connection of Large Language Models, enterprise data, APIs, and internal systems.
- Implementation of RAG, Tool Calling, and agent orchestration, as well as development of reusable tools, prompts, and state and context components.
- Integration of Azure OpenAI and other Foundation Models into Databricks as well as into existing application and data landscapes.
- Development of Python-based agent services and APIs and their integration into existing backend and cloud architectures.
- Use of Databricks Mosaic AI, Model Serving, and MLflow for versioning, evaluation, deployment, and monitoring of models and agents.
- Implementation of Vector Search, Embeddings, and Retrieval Pipelines for controlled access to relevant enterprise data.
- Orchestration of agent, data, and automation processes with Databricks Workflows and defined tool accesses.
- Automation of deployment with GitLab CI/CD and Terraform, including reproducible environments, tests, and traceable releases.
- Development of evaluations, tracing, and monitoring for response quality, groundedness, security, latency, costs, and reliability.
- Establishment of guardrails, authorization concepts, and human-in-the-loop processes for the responsible use of AI Agents.
- Ensuring security and governance with Unity Catalog, Managed Identities, Key Vault, RBAC, Private Endpoints, and Logging.
- Close collaboration with Product Owners, business units, Data, Backend, and Cloud teams, as well as Security and Privacy on productive agent use cases.
- Documentation of agent logic, tool interfaces, data flows, architecture, evaluation procedures, and runbooks for sustainable operation.
Project details
- Start: ab sofort
- Remote: remote
- Location: Köln, Deutschland, DE
- Rate: 95-95 EUR/h
- Languages: German
Indexed by 9am's meta-search. You are a job provider? Claim or manage this listing
Found on FreelancerMap DE