Agentic AI Architect

EnhanceLearning.AI is looking for an Agentic AI Architect to design and shape the architecture of next-generation AI-native and agentic systems.
We are looking for a hands-on architect who understands how LLMs, agents, tools, MCP, memory, context, orchestration, evaluation, observability, security, and infrastructure fit together to create reliable production systems. This is an architecture role, but it is also highly technical. You should be comfortable designing systems, evaluating technologies, building prototypes, and working closely with engineers.
What You'll Do
Design end-to-end architectures for production-grade agentic AI systems.
Define when to use agents, workflows, deterministic services, or human-in-the-loop approaches.
Design single-agent, multi-agent, and hierarchical agent architectures.
Architect agent state, memory, context, retrieval, and knowledge systems.
Design tool ecosystems and integrations using MCP.
Evaluate and select LLMs, models, agent frameworks, orchestration platforms, and AI infrastructure.
Establish patterns for agent reliability, error handling, retries, recovery, and graceful degradation.
Design evaluation frameworks and testing strategies for non-deterministic AI systems.
Define observability and tracing across models, agents, tools, and workflows.
Design security boundaries, permissions, identity, sandboxing, and governance for AI agents.
Identify architectural coupling, scalability concerns, and technology risks.
Build technical prototypes to validate architectural decisions.
Research emerging AI technologies and determine their practical value.
What We're Looking For
Strong background in software architecture and system design.
Significant software engineering experience, preferably 8+ years.
Hands-on experience building LLM, Generative AI, or agentic AI systems.
Strong understanding of distributed systems, APIs, data flows, state management, and reliability.
Strong programming experience in Python and/or TypeScript.
Practical experience designing or implementing AI agents and agentic workflows.
Strong understanding of LLM capabilities, limitations, and failure modes.
Ability to evaluate architectural trade-offs independently.
Valuable Experience
Experience with several of the following is a strong plus:
MCP
LangGraph
LangChain
OpenAI Agents SDK
Multi-agent systems
RAG
Agent memory
Workflow orchestration
Event-driven architectures
Durable execution
LLM evaluation
AI observability
PostgreSQL / Redis
AWS / GCP / Azure
Framework experience is useful, but we care more about understanding the architecture underneath the frameworks.
What You'll Get
Opportunity to shape the architecture of an AI-focused technology platform.
Hands-on exposure to rapidly evolving agentic AI technologies.
Freedom to evaluate models, frameworks, protocols, and infrastructure.
Ownership of architecture decisions and technical direction.
Opportunity to publish technical research and architecture insights.
Opportunity to build reference implementations and practical AI systems.

Work arrangement
No