Staff Research Engineer/Scientist - Conversational AI
Prefer a full account with a password? Sign Up to Apply
Job Description
Staff Research Engineer — Conversational AI
As a Staff AI Research Engineer, you will drive the research and development of agentic AI systems that can reason over real enterprise data, execute complex workflows, and operate reliably at Fortune 500 scale.
You will work at the intersection of applied AI research and systems engineering, exploring new approaches to agent architectures, reasoning, orchestration, and evaluation, while translating promising ideas into robust production capabilities.
- Agent architecture: Research, prototype, and develop advanced agent architectures and design patterns. Explore approaches to planning, reasoning, context management, tool use, memory, and multi-agent collaboration to improve the capabilities of enterprise AI agents.
- Agent orchestration & harness design: Design and evolve the agent execution harness, exploring different orchestration strategies, control flows, and execution models for multi-step agentic workflows. Determine how agents should interact with tools, delegate tasks, manage state, and recover from failures.
- Evaluation & benchmarking: Develop rigorous evaluation methodologies to measure agent capabilities across reasoning, task completion, accuracy, robustness, and efficiency. Build benchmarks, evaluation datasets, and systematic experimentation frameworks to compare architectures, models, and agentic approaches.
- Reliability & agent behavior: Investigate failure modes in agentic systems, including non-deterministic behavior, reasoning errors, tool failures, and cascading execution issues. Develop techniques to improve agent reliability, predictability, and performance in complex enterprise environments.
- Observability & behavioral analysis: Build instrumentation and analytical frameworks that provide visibility into agent behavior, execution trajectories, reasoning patterns, latency, and cost. Use production traces and behavioral data to identify limitations and drive iterative improvements.
- Technical leadership: Set the technical direction for agentic AI research and development. Drive architectural decisions, establish experimentation standards, mentor engineers, and help the team develop a strong foundation in modern agent design and applied AI research.
- Research-to-production engineering: Bridge the gap between research prototypes and production systems. Identify which approaches merit investment, validate them empirically, and work with engineering teams to turn successful experiments into scalable, maintainable enterprise AI capabilities.
Requirements
Department: Engineering, Infrastructure and Operations
Function: Engineering
Experience Level: Not Applicable