AI Backend Engineer (f/m/d)

About the role

paretos builds an AI-driven Decision Intelligence platform that helps enterprises make faster, better data-backed decisions. Its Business Agents can work through large data sets, forecast demand, plan replenishment, and support strategic decisions at leadership level, all without requiring users to write code or have a data science background.

The company describes itself as an inclusive employer with equal opportunity for people of all genders, cultures, and backgrounds. Its culture is shaped by GUNG HO principles: meaningful work, common goals, and mutual support.

This position is based in a hybrid setup, with 1–2 days each week in the Heidelberg office, working closely with Product, Forward Deployed/AI Engineers, and Data Scientists.

What you will do

• Develop the backend foundation that powers the Decision Intelligence platform and brings forecasting, optimization, agent workflows, and the semantic data layer into production.
• Work in an AI-first way using tools such as Claude Code, similar CLI-based assistants, MCPs, and custom skills as part of your day-to-day workflow.
• Build, run, and improve scalable Python backend services with FastAPI and Pydantic to support product workflows around forecasting and replenishment.
• Create robust APIs between the platform, agents, machine learning models, the frontend, and external customer systems such as ERP, BI, and planning tools.
• Strengthen the semantic data layer by modeling in PostgreSQL, building dbt transformations, and handling pipelines, caching, and retrieval.
• Integrate LLM and agent capabilities into production systems, including tool use, function calling, RAG pipelines, evaluation frameworks, guardrails, and deployment.
• Work closely with Product and Forward Deployed Engineers to turn business needs into workable technical solutions and make pragmatic decisions about speed, scalability, and maintainability.
• Prototype quickly, get features into users' hands within days, collect early feedback, and use real usage patterns to decide what should be hardened and rolled out.
• Continuously improve the internal AI development setup, including workflows, hooks, skills, plan/execute/review loops, plugins, tooling, and documentation.

What we are looking for

• 2 to 5 years of professional experience as a Backend, Software, or AI Engineer in a product-oriented environment, ideally in a SaaS or product company; startup experience is welcome.
• Strong technical grounding in Python, FastAPI, Pydantic, API design, asynchronous programming, PostgreSQL, Redis, Docker, and CI/CD.
• A genuine product mindset: you care about user impact, adoption, and measurable outcomes, not only about shipping tasks.
• Hands-on experience with LLMs and agent systems, including prompt design, tool use/function calling, retrieval-augmented generation, evaluation, and production deployment.
• Comfort with cloud infrastructure, preferably AWS, plus familiarity with Kubernetes and Terraform.
• A high sense of ownership, comfort with ambiguity, and motivation to keep growing in backend engineering, agentic engineering, and product thinking.
• Clear communication skills across technical and non-technical audiences, paired with a low-ego mindset.
• Fluent English; native German or C1 German level is a plus.

Benefits

• Company-provided hardware to support your work.
• Personal learning budget of 2,000 EUR per year.
• Flexible working hours.
• Monthly benefit budget of 60 EUR, usable for options such as Urban Sports Club, fuel, public transport, the Deutschlandticket, or gift cards.
• Free snacks and drinks in the office.
• Direct exposure to frontier AI developments and learning sessions with the product team.
• Opportunity to work with a highly skilled, motivated team that enjoys building and collaborating.
• Quarterly Hacky Days focused on creativity and new ideas.
• Regular professional and social team events, both online and in person.
• Dog-friendly office access for members of the paretos Dog Club.
• Office coffee culture, including the chance to show off your barista skills at the Christmas party.

Additional information

AI tools are used to pre-screen applications, but final decisions on interviews are always made by the team, not by software.

paretos encourages applications from candidates who do not meet every requirement, as long as they feel they can contribute and grow in the role.