
Principal AI Agentic Architect
- San Mateo, CA
- Permanent
- Full-time
- Lead the end-to-end architecture of an AI Agentic Platform enabling multi-agent reasoning, orchestration, and real-time dialog management.
- Design a reasoning and planning engine that supports agent task decomposition, tool usage, and contextual decision-making.
- Architect conversation orchestration and dialog management systems for multi-turn, multi-modal interactions across channels (chat, voice, APIs).
- Define multi-agent coordination (A2A) protocols to support collaborative agents and dynamic role allocation.
- Establish agent lifecycle management (build, test, deploy, monitor) as a core platform capability.
- Evaluate and integrate LangChain, LangGraph, LangSmith, and similar frameworks into the platform.
- Define tool orchestration APIs and connectors for agents to interact with enterprise systems.
- Build prompt and knowledge governance frameworks for safe and auditable AI assistant behavior.
- Partner with ML engineers to design LLM reasoning pipelines (chain-of-thought, reflection loops, RAG).
- Architect a cloud-native, multi-tenant SaaS platform (microservices, Kubernetes, event-driven architecture).
- Implement stateful workflows for dialog persistence, memory, and long-horizon planning.
- Define observability and evaluation pipelines (using LangSmith or equivalent) for tracing, testing, and optimizing agents.
- Ensure enterprise security, RBAC, compliance (SOC2, GDPR), and cost-optimized LLM/API usage.
- Serve as a thought leader, mentoring engineers and influencing platform architecture across backend, AI, and UI teams.
- Partner with product and engineering leadership to define the platform roadmap and technical priorities.
- Drive standards for reasoning agents, prompt management, testing, and reliability in production.
- Collaborate with Forward Deployment Engineers (FDEs) to incorporate real-world deployment feedback into the architecture.
- 10+ years of experience in software architecture and backend platform development, including at least 3+ years in AI/LLM-driven systems
- Deep experience in conversational AI systems (orchestration, stateful dialog, context tracking). Experience building autonomous agent ecosystems or multi-agent coordination frameworks.
- Expertise in agentic AI frameworks (LangChain, LangGraph, Autogen, CrewAI, etc.), reasoning/planning agents, and dialog management.
- Proven background in distributed systems, event-driven architecture, microservices, and API design
- Proficiency in Java, Python, plus experience integrating LLM APIs (OpenAI, Anthropic, etc.)
- Strong knowledge of enterprise SaaS architectures: multi-tenancy, RBAC, data governance, and observability
- Experience with workflow orchestration frameworks (Temporal, Airflow) and streaming/ event buses (Kafka, Pub/Sub)
- Familiarity with vector search (Pinecone, ElasticSearch) and retrieval-augmented generation (RAG) pipelines
- Demonstrated leadership in early-stage product/platform development (startup or high-growth team experience preferred)
- Background in conversational platforms (Dialogflow CX, Rasa, LivePerson, or similar)
- Strong knowledge of AI evaluation (LangSmith), reinforcement learning for planning, or reasoning engine design
- Prior exposure to AI governance, safety, and compliance frameworks
- Contributions to open-source LangChain/LangGraph or similar projects are a plus