Staff Applied AI Researcher

Responsibilities

• Set technical direction for agentic reasoning systems and runtime intelligence across ModelMesh™
• Architect infrastructure for researcher augmentation at scale to enable deployment of AI agent fleets
• Advance the science of autonomous reasoning through design and training of routing, verification, and reward models
• Unify perception, retrieval, reasoning, and action into coherent agentic workflows
• Drive research on agent reliability, failure detection, and auditable behavior for regulated environments
• Define evaluation methodologies for task success, decision quality, and robustness
• Influence platform-level architecture regarding model routing, tool use, and governance
• Mentor researchers and maintain hands-on research impact through publications and patents

Requirements

• PhD or MSc in Computer Science, Machine Learning, AI, Robotics, or a related field
• 8+ years in AI/ML research with impact on production systems
• 3+ years building LLM-based or autonomous AI systems
• Deep experience in multi-agent coordination, planning under uncertainty, or tool-using agent systems
• Experience designing evaluation frameworks for non-binary correctness and reliability
• Proficiency in Python with strong software architecture instincts
• Proven technical leadership and experience mentoring researchers

Preferred Qualifications

• Experience with orchestration systems involving dynamic routing and verification loops
• Background in probabilistic modeling, Bayesian inference, or control theory
• Experience with reliability engineering for autonomous AI in regulated environments
• Track record of integrating retrieval, knowledge graphs, and domain models into reliable systems
• Strong publication record and experience moving agent systems from prototype to production

About the Company

Articul8 provides a GenAI platform for regulated industries, combining domain-specific models, autonomous agentic reasoning through ModelMesh™, and reliable model evaluation through LLM-IQ™.