We're building a platform (architect.new) that lets anyone create, deploy, and manage autonomous AI agents with custom skills, tool integrations, and production-ready UIs. You'll work across backend, frontend, and infrastructure — reporting to the Product Architect.

What You'll Do

- Build and maintain the core agent orchestration engine, skill execution, and state management
- Develop the builder UI in Next.js for composing agents and workflows
- Implement real-time communication layers (WebSockets, SSE) for live agent interactions
- Own backend microservices in Python and Go
- Design and optimize MongoDB schemas for multi-tenant, high-throughput workloads
- Manage cloud infrastructure, deployments, and observability on AWS

Required Experience

- Building scalable microservices in Python (FastAPI) and Go
- WebSocket and SSE implementation; diagnosing socket-level issues
- Pub/Sub patterns and event-driven architecture
- Queue management, state persistence, retry and dead-letter handling
- Understanding of latency, networking, and distributed systems performance

Frontend — Next.js

- CSR vs. SSR trade-offs and when to apply each
- Diagnosing hydration issues
- State management across client and server boundaries




- Scalable component architecture
- Deployment, reverse proxying, and build optimization

Infrastructure & Observability

- ECS management and CloudWatch log aggregation
- Cloud networking basics — VPCs, load balancers, security groups
- Managed services — DocumentDB, Redis
- SigNoz or similar observability tools

Version Control

- Git workflows — branching strategies, PR reviews, conflict resolution
- Repository management for microservice codebases

System Design

- HLD: Distributed system architecture, service decomposition, scalability patterns, fault tolerance
- Database Design: Schema modeling (SQL & NoSQL), indexing, query optimization, consistency trade-offs

Nice to Have

- Experience with AI agent frameworks — tool use, skill orchestration, multi-step reasoning
- LLM integration patterns — prompt chaining, function calling, structured outputs
- Firecracker MicroVMs or E2B-style sandboxed execution environments
- Sandbox isolation, security boundaries, and secure code execution
- Agent memory and long-running task orchestration

#J-18808-Ljbffr

📌 Founding Platform Engineer (Centro)
🏢 Lyzr AI
📍 Centro