Senior Software Engineer - AI & Task Mining

Celonis Celonis · Data AI · Munich, Germany · Engineering

Senior Software Engineer to shape the AI and algorithmic core of Celonis' Task Mining solution. The role involves designing and building AI-powered features, advancing the insight engine using LLMs and ML, integrating diverse data sources, developing production-ready Python services, building scalable data pipelines, and implementing robust evaluation metrics. The focus is on turning raw interaction data into structured, actionable insights for enterprise customers.

What you'd actually do

  1. Design and build AI- and algorithm-powered features from scratch — from research and prototyping through to production rollout — across a growing set of problems.
  2. Advance our core insight engine: evolve techniques that combine LLMs, machine learning, and rule-based/algorithmic methods to extract structure from complex, noisy interaction data — keeping results accurate, explainable, and cost-efficient.
  3. Integrate new and diverse data sources into the product, designing the abstractions that let us plug them in cleanly.
  4. Develop and deploy production-ready Python services — clean async APIs, containerized and integrated into a multi-tenant SaaS platform — to serve our AI and data pipelines.
  5. Build and extend robust, scalable data pipelines for ingesting, processing, and transforming large datasets, and design the data models behind them.

Skills

Required

  • 5+ years of practical experience in a Computer Science / Data Science related field, or a PhD in Data Science/AI/ML with 2+ years of practical experience.
  • Hands-on experience applying AI/ML to real problems — including LLMs and Retrieval-Augmented Generation (RAG): writing effective prompts, using vector search, and evaluating outputs for hallucination, schema adherence, and performance (latency/cost).
  • A strong algorithmic foundation and the judgment to choose between AI-based and classical approaches for a given problem.
  • Experience designing, building, and deploying robust, scalable, production-ready Python services and APIs.
  • Strong understanding of Python's async I/O model and why it matters for high-performance, I/O-bound applications.
  • Solid grasp of ETL, data warehouses/lakes, data modeling, and schema design.
  • Experience with containerization (Docker) and CI/CD (e.g. GitHub Actions).

Nice to have

  • PhD in Data Science/AI/ML

What the JD emphasized

  • Hands-on experience applying AI/ML to real problems
  • writing effective prompts
  • using vector search
  • evaluating outputs for hallucination, schema adherence, and performance (latency/cost)
  • strong algorithmic foundation
  • production-ready Python services and APIs
  • Python's async I/O model
  • ETL, data warehouses/lakes, data modeling, and schema design

Other signals

  • design and build AI- and algorithm-powered features from scratch
  • advance our core insight engine: evolve techniques that combine LLMs, machine learning, and rule-based/algorithmic methods
  • hands-on experience applying AI/ML to real problems — including LLMs and Retrieval-Augmented Generation (RAG)