About the Role
Seeking a Senior Backend Developer to build and scale a production-grade LLM agent platform. You’ll work with advanced AI orchestration (LangGraph, LangChain), high-performance async Python services, real-time streaming, multi-provider LLM integrations, and enterprise-grade safety/guardrails. This is not prototype work—the platform is fully tested, observable, documented, and built for scale.
Core Tech Stack
Backend & Services
- Python 3.11+, async/await, mypy (strict typing)
- FastAPI for async APIs (OpenAPI docs)
- Pydantic v2
- UV package manager
- Docker (optimized multi-stage builds)
AI & Orchestration
- LangGraph + LangChain workflows
- Multi-provider LLM support (OpenAI, Anthropic, Azure OpenAI, OpenRouter)
- Dynamic model switching
- Human-in-the-loop (HITL) patterns
- Guardrails AI (PII, toxicity, prompt injection protection)
- MCP for external tool integrations
Infrastructure / DevOps
- PostgreSQL / SQLite / Redis for checkpointer
- Server-Sent Events (SSE) streaming
- OpenTelemetry, LangSmith
- GitHub Actions CI/CD
- Health/readiness checks + structured logging
Key Responsibilities
Architecture & Development
- Build LLM agents using LangGraph
- Implement MVCS architecture (Handler → Service → Gateway)
- Develop internal/MCP tools and multi-provider LLM gateways
- Build streaming endpoints with guardrails
- Optimize async workflows for high concurrency
Code Quality & Testing
- Write unit/integration/E2E tests (pytest)
- Enforce strict typing (mypy) + code standards (Ruff)
- Review PRs, maintain architecture boundaries
- Document APIs, decisions, troubleshooting
Production & Operations
- Deploy horizontally scalable, stateless services
- Implement observability: tracing, logging, metrics
- Improve LLM provider performance/cost
- Implement safety guardrails
- Troubleshoot across FastAPI, LangGraph, and LLM integrations
Collaboration
- Participate in API design
- Mentor junior engineers
- Contribute to documentation & architecture decisions
Required Qualifications
- 5+ years Python experience
- Strong async/await, FastAPI (or similar async frameworks)
- Experience with LLM APIs
- Docker + cloud deployment
- Strong understanding of REST, layered architecture, type systems
- Expertise in testing, CI/CD, security, and performance tuning
- Experience with streaming (SSE/WebSockets) and scalable systems
Preferred Qualifications
- Experience with LangChain/LangGraph
- LangSmith, Guardrails AI, MCP familiarity
- Python tooling: UV, Ruff, Pydantic Settings
- OpenTelemetry, distributed tracing
- Redis/PostgreSQL experience
- Knowledge of rate limiting, resilience patterns
- Experience with regulated domains / AI safety concerns
What Makes This Role Unique
- Production-grade LLM orchestration (not prototypes)
- Multi-provider, per-request model switching
- HITL workflows + real-time validation pipelines
- MVCS architecture with strict boundaries
- Strong quality, testing, safety, and observability culture
