About the project

• We’re looking for a QA Tester with strong AI literacy and data validation skills to test how AI agents behave and ensure that their outputs are stored accurately and reliably in backend systems. This is a manual, exploratory testing role — ideal for someone who can combine curiosity about AI behavior with practical knowledge of data flows, relational databases, and result traceability.
• You won’t be writing automation scripts, but you will need to understand how AI agents operate, how their outputs are used, and how to verify correctness across both the UI and backend.

Your responsibilities

• Manually test AI-driven workflows that generate content, complete tasks, or make decisions
• Assess AI behavior by checking for: Consistency and repeatability; Hallucinations, inaccuracy, or bias; Relevance and task alignment
• Evaluate data integrity: Trace AI-generated data from interface to backend; Use SQL to validate how outputs are stored, structured, or logged; Compare AI intent/output to resulting records in the database
• Reproduce and report subtle, fuzzy, or probabilistic issues with structured documentation
• Collaborate with engineers, AI designers, and product owners to define quality criteria across system layers

Our requirements

• 2+ years of manual QA experience, ideally in exploratory or context-driven testing environments
• Practical understanding of LLMs and AI tools (ChatGPT, Claude, etc.)
• Basic to intermediate knowledge of SQL (joins, filters, aggregations, subqueries)
• Experience validating data pipelines, audit logs, or relational integrity
• Able to detect both UI anomalies and backend data discrepancies
• Clear written and verbal communication for reporting behavior-based bugs
• Familiarity with testing non-deterministic or AI-powered systems
• English level min.B2

Optional

• Understanding of prompt engineering and how LLM behavior can shift with input changes
• Familiarity with AI agent architectures (e.g., LangChain, ReAct, RAG systems)
• Experience working with BI tools (e.g., Metabase, Redash) or data validation frameworks
• Background in content moderation, safety testing, or AI/UX evaluation

Technologies we use

This is how we work on a project

• functional tests

• integration tests

• performance tests

• regression tests

• unit tests

• manual tests

What we offer

• Long-term cooperation

• Work-life balance: flexible working hours and a hybrid model (2 times per week in the office), usually it is Wednesday and Thursday

• Semi-annual evaluation meetings and a clearly defined career path and salary forecast

• A friendly Buddy to guide you through the onboarding and further training and career opportunities

• Training possibilities for personal development after completing the onboarding period

• An open and casual company culture filled with internal events

• The opportunity to quickly develop professionally on projects in Europe and all over the world