Full-Stack Software Engineer, Core Platform

About the Role

Lumbra is building Nebula, an agentic harness that transforms how analysts work with information, turning days of manual synthesis into minutes of AI-augmented insight. We're looking for a full-stack engineer to own features end to end across our Python backend and React frontend, shipping capability that puts powerful AI tooling directly in the hands of users.

What You'll Own

• Design and build backend API endpoints in FastAPI backed by PostgreSQL and SQLAlchemy to power agent configuration, workflow orchestration, and evaluation pipelines. You need strong Python skills and comfort with relational data modeling, async patterns, and layered service architectures.

• Implement interactive frontend features in React with TypeScript, including conversation interfaces, workflow visualization, and knowledge graph exploration. Comfort with modern React patterns and state management is essential.

• Build and maintain durable workflow orchestration for long-running agent execution, including delegation, monitoring, and timeout propagation. Experience with distributed task orchestration systems is a strong plus.

• Own features across the full stack, from database migrations through API contracts to UI state management. You need to be comfortable reasoning about data flow through every layer of the system.

• Bring sound software engineering fundamentals to a fast-moving codebase: clean abstractions, well-defined interfaces, and architecture that survives rapid iteration. In an era of AI-assisted development and constant change, we value engineers who write code that is easy to understand, refactor, and extend.

• Write contract-driven tests across both backend and frontend, ensuring spec coverage rather than implementation coupling. You'll follow a test-first development approach.

• Design testing strategies for non-deterministic AI components, including metamorphic testing, property-based testing, and regression suites that catch capability regressions without overfitting to specific model outputs.

• Contribute to our Python and TypeScript client SDKs, ensuring programmatic consumers have ergonomic, well-typed access to platform capabilities.

• Own performance across the stack: API response times, query efficiency, frontend render performance, and end-to-end latency through streaming agent workflows. You profile before you guess, and you treat slowness as seriously as a broken feature.

• Debug and resolve issues that cross service boundaries, using local development environments and distributed tracing to isolate problems quickly.

Preferred Qualifications

• Experience with agentic AI systems, including agent harness engineering, context engineering, or building applications on top of LLM APIs

• Familiarity with graph databases or vector similarity search for knowledge representation and semantic retrieval

• Prior work with the Model Context Protocol (MCP) or similar tool-use frameworks

• Experience with identity management systems (OAuth2/OIDC)

• Contributions to open-source developer tooling or frameworks

• Background in intelligence analysis, OSINT, or analytical workflows that inform product intuition

Experience with cloud infrastructure (GCP or equivalent): managed Kubernetes, managed databases, secrets management

Benefits

• Comprehensive medical, dental, and vision plans

• Premiums 100% covered by Lumbra for all employees

• Exceptionally low premiums for spouses and dependents

• Basic life insurance and disability 100% covered for all employees by Lumbra

• Option to purchase additional life insurance available

• Take the time off that you need, when you need it' paid time off, not accrual based

• Generous company holiday calendar including a holiday shutdown in December

• Supportive leave of absence program including time off for military service, medical events, and parental leave

• Full 401(k) retirement plan for all full-time eligible employees

• Company-funded commuter benefits

• Free access to on-site gym at office