About The Position

Research and develop multi-agent orchestration and decision-policy algorithms, including inference-time agent routing, task decomposition, and coordination across open-weight and proprietary LLM and Transformer-based models. Develop and scale vector-based retrieval and retrieval-augmented generation (RAG) systems for high-volume agentic reasoning across multiple AI applications. Research novel training algorithms to specialize and align multi-agent LLM systems, and productionize them as platform-level training capabilities. Design and scale platform-level evaluation frameworks for LLM-driven generation and inference systems. Create graph-based knowledge representation and reasoning algorithms over semantic knowledge graphs to support multi-agent LLM workflows. Develop distributed multi-agent AI platform infrastructure on AWS, including application pipelines developed in Swift. Research and implement safety and robustness methods for multi-agent LLM systems, including defenses and evaluation for agentic workflows operating against building protocol networks.

Requirements

  • Master's degree in Computer Science.
  • Two years of experience as an LLM Systems Scientist or AI Engineer.
  • Proficiency with Python, PyTorch, TensorFlow, NumPy, pandas, DGL, SQL, Git, AWS, Swift, LLMs, Transformer-based systems, RAG, vector-based retrieval, and graph neural networks.

Responsibilities

  • Research and develop multi-agent orchestration and decision-policy algorithms, including inference-time agent routing, task decomposition, and coordination across open-weight and proprietary LLM and Transformer-based models.
  • Develop and scale vector-based retrieval and RAG systems, including vector index optimization, retrieval-latency tuning, hybrid dense and sparse search across SQL datastores, and embedding-pipeline throughput.
  • Research and develop novel training algorithms, including contrastive learning, to specialize and align multi-agent LLM systems; productionize them as platform-level training capabilities managed under Git-based version control and CI/CD.
  • Design and scale a platform-level evaluation framework, including statistical benchmarking pipelines, automated regression suites, and quality and reliability dashboards for LLM-driven generation and inference systems.
  • Create graph-based knowledge representation and reasoning algorithms using graph neural networks over semantic knowledge graphs, supporting graph generation, node and edge inference, and graph-grounded retrieval.
  • Develop distributed multi-agent AI platform infrastructure on AWS for production agent deployments serving multiple product lines, including application pipelines developed in Swift.
  • Research and implement safety and robustness methods for multi-agent LLM systems, including adversarial and prompt-injection robustness, data-exfiltration defenses, agent action sandboxing, and behavioral safety evaluation and red-teaming.
© 2026 Teal Labs, Inc
Privacy PolicyTerms of Service