Job Title: Research Assistant – AI Agents & LLM PrototypingLocation: London, UK- Hybrid- 3 Days onsiteDuration: Until Dec 2025- possible extensionRole Overview: We are looking for a Research Assistant with 2+ years of experience in prototyping and testing AI agents or large language models (LLMs). You will design test prompts, experiment with prompt engineering, and debug AI agent tool calls within a Python/PHP software stack. You’ll also help create internal benchmarks to evaluate AI agent performance.Key Responsibilities:Create and refine test prompts to guide AI agents toward desired behavior.Implement and troubleshoot AI agent tool calls in a Python/PHP environment.Develop high-quality prompts to build internal evaluation benchmarks for AI agents.Test AI agents to assess their ability to perform tasks such as ordering, scheduling, or cancelling meetings.Analyze test outcomes, identify issues, and communicate findings for continuous improvement.Navigate and understand the Python codebase to correlate test results with underlying code.Improving and testing existing AI agents.Identifying where agents perform well and where they fail.Prompt testing and tuning to optimize agent responses.Flagging test results as pass/fail based on expected behavior.Technical Skills & Qualifications:Bachelor’s degree in Computer Science, Computer Engineering, or related field (or equivalent experience).2+ years of experience in research or prototyping within machine learning, deep learning, or natural language processing.Proficient in Python, with the ability to understand and work with a codebase containing multiple interrelated files.Experience with AI agent frameworks such as Model Context Protocol (MCP) and LangChain.Familiarity with architectural patterns of large-scale software systems.Basic knowledge of SQL and data analysis is a plus.Experience using source control systems (e.g., Git).Experience in generative AI and LLM research preferred.Essential Skills:Python programmingGenerative AI knowledgeUnderstanding of MCP servers and LangChain frameworks
