IT Craft is looking for an experienced AI Developer with hands-on expertise in LLM evaluation and automated testing to join our team. In this role, you will take ownership of designing and implementing an evaluation framework for a deployed AI assistant running on AWS Bedrock, with infrastructure support from the internal DevOps team.
Key Responsibilities:
- Design and implement automated end-to-end and regression testing infrastructure for an LLM-based assistant;
- Build an evaluation harness that runs a structured set of prompt scenarios against the deployed service;
- Implement automated checks for response correctness, tool usage, and guardrail behavior;
- Generate clear pass/fail results based on agreed evaluation baselines;
- Collaborate with the internal DevOps team to integrate evaluation runs into CI/CD processes and identify quality and behavior regressions before and after deployments;
- Analyze evaluation results and continuously improve test coverage and reliability;
- Plan and organize the implementation independently, define follow-up tasks, and coordinate with internal specialists through regular synchronization meetings.
Required Skills:
- Hands-on experience with LLM evaluation tools such as Promptfoo, Langfuse, Arize, or comparable solutions;
- Practical experience designing or implementing automated evaluation and regression testing for LLM-based applications;
- Understanding of approaches used to evaluate non-deterministic LLM responses, tool calls, and guardrail behavior;
- Solid experience with AWS and CI/CD;
- Ability to design a solution from scratch and deliver it independently;
- Familiarity with Kotlin and Spring would be a plus;
- Strong written and spoken English (Upper-Intermediate+ / Advanced).