Responsibilities
- Create robust architectures for AI agent workflows, including planning cycles, tool integration, memory systems, retrieval mechanisms, and human oversight checkpoints
- Assess, integrate, and fine-tune foundational models and large language model APIs tailored to specific enterprise applications and data requirements
- Establish production-grade standards for agent dependability, monitoring, and handling of failure scenarios
- Work closely with field engineers to convert successful client implementations into standardized, reusable platform features
- Develop internal tools and evaluation frameworks to measure agent performance, hallucination rates, and task success
- Make well-documented, strategic architectural choices while maintaining awareness of emerging technologies to guide adoption
Requirements
- 6 to 10 years of experience building AI or data systems that operate reliably at scale in production, not just prototypes
- Extensive practical knowledge of multi-agent systems, including context window handling, memory management, dependency structures, and real-world failure points
- Proficiency in Python and experience with agent frameworks such as LangChain, LlamaIndex, AutoGen, or equivalent; or a well-reasoned alternative approach
- Hands-on experience implementing RAG systems, vector databases, and managing context windows in live environments
- Proven track record deploying LLM-based systems in enterprise settings with attention to data security, access controls, and audit trails
Nice to Have
- Exposure to machine learning research practices, including fine-tuning, reinforcement learning from human feedback (RLHF), and evaluation methodologies
- Experience with AI deployments in regulated sectors such as finance, insurance, or healthcare
- Familiarity with blockchain data systems or institutional cryptocurrency infrastructure
Compensation
Competitive base salary and meaningful early-stage equity
