Note: The job is a remote job and is open to candidates in USA. BigBear.ai provides AI-powered decision intelligence solutions for national security, supply chain management, and digital identity. The AI Application Engineer will develop and maintain AI applications and a config-driven decision-support platform, owning end-to-end features across backend APIs, databases, LLM/RAG pipelines, and React-based user interfaces.
Responsibilities
- Build and maintain full-stack features: FastAPI services, PostgreSQL schema/migrations, and React/TypeScript UI in an Nx monorepo
- Design, implement, and maintain LLM/RAG pipelines using LangChain (Python) or LangChain4J (Java)—retrieval, embeddings, tool/agent orchestration, and prompt workflows
- Integrate LLM providers (e.g. OpenAI, Anthropic, Ask Sage), document ingestion, semantic search, and chat/report features
- Work with vector stores (e.g. PostgreSQL (pgvector), Weaviate), Redis, and Keycloak auth; implement role-based access and secure API patterns
- Connect to external systems via MCP, REST/PostgREST, and related data adapters
- Write and maintain tests (pytest, Vitest) and participate in CI/CD (Docker, GitHub Actions)
- Debug production issues, improve reliability/performance, and keep dependencies and security patches current
- Use AI-assisted development tools (e.g. Cursor) effectively: prompt well, review generated code critically, and ship production-quality changes at a steady pace
Skills
- All applicants must currently reside in the United States
- All applicants must reside in the National Capital Region
- 3+ years full-stack development with strong Python and React/TypeScript
- Must be able to obtain and maintain a security clearance
- Hands-on LLM/RAG experience building production features with LangChain or LangChain4J (chains, retrievers, agents/tools, embeddings, vector stores)
- Experience building and consuming REST APIs (FastAPI or similar)
- Solid PostgreSQL skills: schema design, SQL, migrations (pgvector a plus)
- Comfortable with Docker, local dev environments, and basic deployment workflows
- Strong debugging, testing, and code review habits
- Ability to work independently across backend, frontend, database, and AI layers
- Proven comfort using AI coding assistants as a daily workflow—not as a substitute for understanding the codebase
- Dash/Plotly, Flask, or legacy-to-modern UI migration experience
- Keycloak/OAuth, Redis sessions
- Nx monorepos, TanStack Router/Query, Tailwind CSS
- MCP protocol, Ask Sage, or similar enterprise LLM platforms
- Government/defense or regulated-environment experience
Company Overview