Stelle ist zu besetzen ab: 01.07.2026

Stellenbeschreibung

Start : June–July (flexible) | Location: [Karlsruhe/Hybrid]

Thesis option : possible MSc About us We are a research-driven spin-off building an on-premise AI platform that helps organizations use their internal knowledge safely. Our focus is governance by design: access control, provenance/traceability, privacy-by-default, review/approval workflows, and audit logs — so AI can be used in regulated environments without losing data sovereignty. What you will build As an early team member, you will help deliver pilot-ready software and iterate quickly with real users. Typical tasks include: Backend services & APIs: ingestion, indexing, retrieval/search, integrations RAG & inference integration: connect OpenAI-compatible endpoints (local or remote), build retrieval pipelines, and expose them via stable APIs Governance features: implement review/approval flows, logging/audit trails, and access-aware retrieval Pilot deployments: make the system reproducible (Docker), run it in customer-like environments, troubleshoot and improve reliability Core skills (we value hands-on builders) Backend (strongly preferred) Python (FastAPI preferred), clean API design, pragmatic testing PostgreSQL (SQL, data modeling) React (TypeScript preferred) and basic UI engineering for configuration/admin flows Nice to have (helpful, not required) Vector search / retrieval (pgvector) LLM serving or integration (vLLM, Ollama, OpenAI-compatible APIs) Document processing (Apache Tika / PyMuPDF) Docker, basic CI/CD, observability (logs/metrics) Security fundamentals: OIDC, TLS, secrets handling Interest in protocol-based integrations (federation, policy enforcement) How we work We don’t expect you to know everything from day one. We care more about learning speed, ownership, and shipping. You will work closely with the founders on architecture and pilots — short feedback loops, real constraints, real users. What we offer A hands-on role in an early-stage AI spin-off with direct impact Practical experience building an on-prem AI platform under governance and security constraints Fast learning curve, responsibility, and short decision paths Thesis-friendly environment (MSc), if desired

Art der Anstellung

• Studentische Hilfskraft / Werkstudent:in

Gewünschtes Studium

• Ingenieurwissenschaften
  Informatik
  Information System Engineering and Management
  Computer Science
• Naturwissenschaften und Technik
  Computational and Data Sience

Gesuchter Karrierestatus

• Studierende:r

Arbeitsregion

• Karlsruhe und Umgebung

Unternehmensbereich

• Forschung

Unternehmensbranche

• Wissenschaft & Forschung

Sprache am Arbeitsplatz

• Deutsch und Englisch

Art des Unternehmens

• Wissenschaftliche Einrichtung