Building agentic LLMs (ReAct/AutoGPT) and tool-using autonomous agents — heavy vibe-coding and agent design.
About the Role
Research and develop agentic AI frameworks that enable LLMs to plan, use tools, and reason through multi-step goals to solve complex enterprise problems. Lead design and implementation of tool-use, long-term memory, and multi-agent/swarm intelligence systems using advanced model development frameworks.
Job Description
Role
United Force Technologies is seeking an Agentic AI Researcher to research and develop agentic frameworks that enable large language models (LLMs) to plan, use tools, and reason through multi-step goals to solve complex enterprise problems. This is a full-time role within the Data & AI department.
Key Responsibilities
- Research and design agentic architectures and specialized agent systems (single- and multi-agent).
- Enable LLMs to perform tool usage via function calling and to maintain long-term memory for extended tasks.
- Develop and evaluate approaches such as ReAct and AutoGPT-style agent behaviors.
- Build multi-agent systems and swarm intelligence solutions for complex enterprise workflows.
- Apply symbolic logic and probabilistic reasoning to improve agent decision-making.
- Implement and prototype custom model components using deep learning frameworks.
Requirements
- PhD in AI/ML with focus on Reinforcement Learning, Planning, or a closely related area.
- Deep understanding of ReAct, AutoGPT, and specialized agent architectures.
- Experience with tool-use (function calling) and designing long-term memory systems for AI.
- Proven ability to build multi-agent systems and swarm intelligence.
- Strong background in symbolic logic and probabilistic reasoning.
- Track record of groundbreaking research or innovative product builds.
- Mastery of PyTorch or JAX for custom model development.
Details
- Employment type: Full-time
- Department: Data & AI
- Salary: $180,000 - $250,000 (as listed)