AI / Backend Engineering

kalkulai
kalkulai

Software Engineering, Data Science

Munich, Germany

Posted on Jul 31, 2026

Multi-agent infrastructure. Already built: a working multi-agent setup and a quote schema. With our team you extend and harden the system for production, live at real pilot customers. Seven pilots already run on it. You never start from zero, you build on real progress and help shape an AI startup from the inside.

What you’ll do

  • Extend the KAI orchestrator and its sub-agents that run a company’s back office.
  • Build on the existing backend and APIs.
  • Ship new product features to production, from the quote schema 2.0 and planning logic to KAI Voice and the e-invoice and DATEV integration.
  • Strengthen the knowledge graph and structured outputs so KAI answers reliably, with no hallucinations.
  • Harden the system for production: error handling, retries and observability.
  • Write evals and regression tests so the agent’s quality never silently degrades.

What you bring

  • CS Master’s (ideally TUM CIT) or similar. Solid coding skills, please share your GitHub.
  • Independent and product-minded: you want things running in production, not just in a notebook.
  • AI-assisted development is a must: Claude, Codex, GPT, Gemini or similar, it is how we work every day.
  • Experience with LLMs, RAG or agent frameworks (LangChain or similar).
  • You are comfortable owning a feature end to end: design, ship, monitor.
  • English or German fluent.
Stack
PythonLangChainpgvectorFastAPISupabasePostgreSQLDocker

Why KalkulAI

  • Radical ownership from day one. No fixed job description, you own outcomes.
  • A live product with real pilot customers, not a lab hypothesis.
  • You never start from zero: you build on real, existing progress.
  • AI-native: Claude, Codex, GPT and Gemini are part of how we work every day.
  • You sit side by side with the founding team.
  • Honest deal: this is unpaid, but you gain real ownership and startup experience most roles never offer.
  • Backed by EXIST, TU Munich and UnternehmerTUM.

How to apply

  1. 1 Apply with your CV, GitHub and a short note on which role and why, to leon@kalkulai.de. Multiple applications are welcome.
  2. 2 A first call with the team, 20 to 30 minutes. Depending on the role, a small case or take-home task can follow.
  3. 3 After a yes, we set up the format together: as an unpaid internship, or we handle the TUM IDP registration with a supervising chair.
Backed by
See more open positions at kalkulai