Principal Engineer

We're looking for a Principal Engineer in Applied AI to define the architecture and engineering standards for Lighthouse, our enterprise AI platform, and the production AI systems built on it.

Responsibilities:

• Own the architecture across the AI platform, context layer, data adapters, workflow/agent runtime, evaluation, and production operations.
• Design scalable systems for enterprise AI, RAG/context, model & tool orchestration, workflows, and long-running state.
• Define architecture for security, authorization, multi-tenancy, tenant isolation, auditability, and customer-hosted deployments.
• Establish standards for AI evaluation, observability, tracing, grounding, reliability, latency, cost, and model/tool fallbacks.
• Personally design, implement, and review high-risk architectural, security, performance, and production components.
• Mentor engineers and drive technical decisions across product, engineering, and customer-facing teams.

Requirements:

• 8+ years of experience building production software with principal-level technical leadership.
• Strong hands-on expertise in Python, API/service design, relational data modeling, asynchronous and event-driven systems, and cloud-native engineering.
• Proven experience with distributed systems, enterprise platforms, security, authorization, multi-tenancy, observability, and failure recovery.
• Experience taking AI/ML systems from prototype to production, including LLMs, RAG/retrieval, tool use, structured outputs, evaluation, guardrails, and production debugging.
• Strong architectural judgment around system boundaries, scalability, consistency, security, data ownership, and build-vs-buy decisions.
• Required tech stacks: Python, FastAPI, Pydantic, PostgreSQL, Redis, Kafka/Queues, MCP, Vector/Graph Stores, Docker, Kubernetes, AWS/Azure/GCP.

Good to Have:

• Experience with enterprise metadata, semantic/knowledge-graph systems, workflow engines, data platforms, developer platforms, multi-tenant SaaS, open-source platform extensions, or regulated/customer-controlled environments.