Jobs at Apollo Research

Design, run, and automate pre-deployment evaluations of frontier AI models. Analyze large model transcripts to surface behaviors, build scalable evals, collaborate with AI labs, and improve evaluation pipelines and tooling.

Build and maintain backend tools for AGI safety research: eval libraries, orchestration for parallel agentic evaluations, LLM proxy and telemetry, CI optimizations, data warehousing, and researcher-facing tooling. Lead feature development, collaborate with researchers, and promote good software design and reliability.

Build and maintain full-stack features for an AI-agent safety product: responsive React front-ends, Python backends, REST APIs, scalable data pipelines, monitoring/observability, and secure integrations. Collaborate with researchers, participate in design and code reviews, and deliver production-ready tooling and visualizations for AI safety monitoring.

Conduct independent, governance-focused research on AI scheming and loss-of-control risks; produce reports, papers, legal analyses; collaborate with technical teams, policymakers, and external researchers; mentor junior researchers and shape team strategy.

Generalist People Ops role focused on hiring, HR systems, process improvement, talent outreach (events, hackathons, conferences), and onboarding. Use AI tools to automate workflows and support a technical workforce of researchers and engineers. Drive substantial headcount growth and build strong candidate relationships across UK, US, and Europe.

Build scalable backend systems to ingest, process, and store large volumes of AI agent logs in real time. Design APIs, authentication, rate limiting, webhooks, and SIEM integrations. Implement data pipelines, storage for structured and unstructured data, caching, retention policies, monitoring, observability, and reliability strategies. Collaborate with researchers and frontend engineers to translate prototypes into production-ready, secure, low-latency monitoring products.

Build full-stack tools for AGI safety research: evaluation IDE, LLM-powered search, comparison views, streaming results, collaborative log editing, pipelines, agents, and telemetry. Collaborate with researchers and product to deliver robust, extensible software.

Conduct empirical research into AI "scheming": design and run RL experiments on LLMs, build model organisms, develop scalable evaluation techniques, study scaling laws and AI cognition, and collaborate with partner labs to influence deployment and mitigation strategies.

Support day-to-day finance operations: run Procure-to-Pay and Order-to-Cash cycles, manage supplier and customer invoicing, prepare payments, follow up overdue balances, reconcile multi-entity multi-currency bank accounts weekly, and ensure accurate transaction recording in Xero.