1 month ago
London, United KingdomSenior
Responsibilities
- Build and deploy AI agents that interpret, prioritize, and act on geopolitical and financial signals.
- Develop tools connecting agents to web access, semantic and hybrid retrieval, structured-data queries, internal APIs, and knowledge graphs.
- Design, implement, and evaluate agentic workflows and multi-agent systems.
- Define metrics and evaluation frameworks for retrieval quality, agent performance, factual accuracy, and business relevance.
- Collaborate with Research, Product, and Engineering to prototype, test, and refine capabilities.
- Take solutions from prototype to production through API development, GitHub Actions CI/CD pipelines, and Kubernetes deployments on AWS.
- Own systems across architecture, implementation, deployment, adoption, monitoring, maintenance, and continuous improvement.
Requirements
- 5+ years of professional experience in AI engineering, data science, software engineering, data engineering, or a closely related field.
- Strong proficiency in Python and SQL.
- Production experience building and operating RAG pipelines and agentic systems.
- Hands-on expertise in BM25, vector search, hybrid retrieval, reranking, and retrieval evaluation.
- Experience designing APIs and deploying containerized services with Docker and Kubernetes.
- Experience rapidly prototyping agentic workflows and making pragmatic decisions about hardening, scaling, or discarding systems.
- Strong product judgment, ownership, roadmap definition, and ability to execute independently.
- Interest in or experience with global security, financial crime, sanctions, export controls, or related domains.
- Experience with graph-aware retrieval, graph databases, or knowledge-graph integration is a bonus.
Benefits
- Fully sponsored private health insurance.
- Pension plan with a 3% employer contribution.
- Annual leave plus bank holidays.
- Annual remuneration of £105,000–£120,000 plus a discretionary annual bonus, paid monthly.
- Full-time London-based role with a hybrid arrangement of three days per week.
- The role has a 180-day probationary period.
Tech Stack
Categories
