Senior AI/ML Engineer

Agentic AI (core)

• Design and build multi-step agents that plan, call tools and APIs, keep state, recover from errors, and hand off to humans for approval, running inside enterprise systems like SAP and Salesforce.
• Work across the modern AI stack: RAG, MCP servers and clients, multi-agent setups, and new models and techniques as they emerge. Quickly test new tools and bring the useful ones into production.
• Own prompt design and evals: measure agent quality, catch regressions, and decide with data whether a change ships.
• Build guardrails: input and output checks, prompt-injection defence, and logging for audits.

Classic ML (supporting)

• Use classic ML models (classification, anomaly detection) as building blocks inside agent workflows.
• Prepare data and measure model results with the right metrics.

Delivery and team

• Take features from idea to production, with clean code, tests and docs.
• Join client calls when needed to understand requirements and demo results.
• Review code and guide 1–2 junior engineers.

Must-have skills

• At least one AI/ML project you built that is live in production.
• Strong Python, plus pandas, NumPy and scikit-learn.
• At least 6 months of hands-on work with a deep learning framework (PyTorch or TensorFlow).
• Built at least one LLM app (Claude, OpenAI or Bedrock APIs) with RAG or tool calling, and have hands-on experience with agent frameworks (LangGraph, LlamaIndex, Claude Agent SDK) and MCP.
• Comfortable with SQL, REST APIs and Git.
• Has deployed models or services using Docker on a cloud (AWS or Azure).
• Can explain technical work clearly to non-technical people.

Nice to have

• Vector databases such as pgvector, Pinecone or Qdrant.
• Experience with SAP, Salesforce, ServiceNow or other enterprise systems.
• Workflow tools like n8n.
• Agent observability and LLMOps tools such as Langfuse or LangSmith.
• Fine-tuning small models or working with open-source LLMs.