Junior Software Engineer (Remote - AI Agent Evaluation)
Company: Leading AI Operations Platform / Tech Services Partner
Location: United Kingdom (Fully Remote)
Employment Type: Full-time / Flexible Contract (Hourly)
Seniority Level: Junior-Mid level
Estimated Salary: USD $30 - $50 per hour (~£4,000 - £6,800 per month based on full-time hours, issued weekly)
Industry: Artificial Intelligence / Software Engineering / LLM Training
Company Overview
Our partner is a premier global AI engineering platform collaborating with the world's leading AI organizations to pioneer the next generation of autonomous AI agents.
By integrating advanced human-in-the-loop technical expertise, we stress-test, evaluate, and train Large Language Models (LLMs) across complex, multi-step architectural workflows. We help shape agentic systems spanning health, education, enterprise software, and real-world systems through high-density technical feedback and advanced codebase auditing.
Key Responsibilities
AI Agent Evaluation & Auditing: Write definitive evaluation rubrics containing objective pass/fail criteria to assess autonomous AI agents across multiple frontier LLM models.
Trace Debugging & Analysis: Deep-dive into execution logs and agent traces to pinpoint subtle failure patterns, loop conditions, and execution bottlenecks in multi-turn system interactions.
Adversarial Stress Testing: Conduct robust stress tests against autonomous agents to uncover vulnerabilities related to edge cases, prompt injection, authority escalation, and tool misuse.
Production Architecture Assessment: Review production-grade modular software architectures to analyze system behaviors, state-tracking efficiency, and API call validation.
High-Density Technical Feedback: Generate comprehensive, highly analytical technical write-ups and documentation to train and optimize LLMs for complex developer workflows.
Experience & Qualifications
- 1–3 years of professional experience in backend software engineering, AI automation pipelines, or complex digital systems integration.
- Proven ability to build and maintain production-grade software applying modular separation principles (e.g., separating data parsing, business logic, and reporting layers).
- Hands-on experience building for live, non-mocked deployment environments and effectively handling complex multi-turn system interactions.
- High autonomy with the ability to manage flexible task assignments independently in a fully remote environment.
- Required Setup: Must complete the onboarding assessment, verify identity/credentials, and manage tasks through the platform interface.
Technical Skills Required
Core Languages: Strong command of at least two major backend programming languages, such as Python, JavaScript/TypeScript, Go, or Java.
Databases & Tooling: Practical fluency working with SQL databases, query optimization, and consuming production-grade REST/GraphQL APIs.
Nice to Have: Technical experience integrating autonomous frameworks with live external tools (e.g., Supabase, Gmail API), managing persistent state tracking, or mitigating indirect prompt injection and privacy leaks.
How to Apply
To join our global remote engineering squad and train the future of agentic AI systems, please upload your developer resume, verified technical identification, and code repository portfolio through our automated onboarding assessment portal.