Agentic workflows are transforming the way software systems operate by enabling AI agents to reason, plan, use tools, and collaborate to solve complex tasks. As these workflows become increasingly sophisticated, ensuring their reliability and consistency across successive software updates is a key challenge. Changes to prompts, models, tools, or orchestration logic can affect an agent’s behavior and lead to unexpected regressions.

You will join a team of AI specialists and software developers working on agentic AI technologies, contributing to the development of robust testing solutions that ensure the quality, reliability, and continuous improvement of AI-driven workflows.

As an AI Engineer Intern (m/f/d), you will contribute to the development of a non-regression testing system for agentic workflows. Your mission will be to design and implement automated testing approaches that evaluate agent behavior, detect regressions, and ensure that changes to AI components do not compromise the expected performance of existing workflows.

Responsibilities

Design and develop an automated non-regression testing framework for agentic workflows.
Define test scenarios, evaluation criteria, and performance metrics to assess the reliability and consistency of AI agents.
Implement automated tests to detect behavioral regressions following changes to prompts, language models, tools, or workflow orchestration.
Investigate approaches for evaluating non-deterministic AI systems, including output comparison, semantic evaluation, and LLM-based judging.
Integrate testing capabilities into continuous integration (CI) pipelines to enable systematic and reproducible evaluation.
Analyze test results, identify failure patterns, and contribute to improving the robustness and quality of agentic workflows.
Collaborate with AI researchers, software developers, and QA engineers to establish best practices for testing and validating agentic systems.

Strong programming skills in Python and good software engineering practices.
Familiarity with generative AI, large language models (LLMs), and prompt engineering.
Understanding of automated testing principles, including unit testing, integration testing, and regression testing.
Knowledge of Git and continuous integration tools is a plus.
Experience with AI agent frameworks, tool calling, or workflow orchestration is a plus.
Familiarity with LLM evaluation techniques, benchmarking, and automated quality assessment is a plus.
Analytical mindset, problem-solving skills, and a strong focus on software quality and reliability.
Ability to work independently while collaborating effectively within a multidisciplinary team.

Intern (m/f/d) AI Development Engineer

Mit deiner E-Mail starten