AI Systems Builder – Multi-Agent Orchestration, Backend Automation

Role Overview

• Build and ship multi-agent orchestrator systems using Claude Code / agent frameworks
• Design agent roles, workflows, and memory structures for real business operations
• Implement coordination logic between agents (task delegation, state passing, retries)
• Ensure systems are production-ready (not demos or notebooks)
• Continuously refine agent behavior based on real-world execution outcomes
• Convert business workflows into automated AI-driven pipelines
• Build standalone agent SDKs and reusable automation modules
• Design systems that execute multi-step business processes without human intervention
• Ensure workflows are reliable, observable, and maintainable in production environments
• Optimize for failure recovery, not just successful execution paths
• Design and implement PostgreSQL schemas for agent systems and workflows
• Build Supabase-backed applications powering AI agents and automations
• Ensure data integrity across multi-agent operations
• Structure databases for real-time agent execution and state tracking
• Debug and optimize backend performance when workflows break
• Build and maintain production API integrations across business tools
• Connect AI agents to external systems (CRMs, property tools, communication tools)
• Handle authentication, rate limits, retries, and failure states
• Ensure integrations are stable and production-safe, not experimental scripts
• Debug end-to-end system flows when integrations fail
• Take systems from architecture → build → deployment → production use
• Package AI systems into usable tools or SDK-like deliverables
• Ensure systems are deployable, maintainable, and configurable
• Write clean, production-grade code (not prototypes or notebooks)
• Maintain and improve systems after deployment based on real usage
• Identify missing requirements without being told
• Break down ambiguous system requests into buildable components
• Choose implementation paths independently
• Debug production issues without escalation dependency
• Maintain momentum without waiting for structured assignments

Requirements

• Proven experience building and shipping multi-agent systems in production
• Hands-on experience with OpenClaw, Hermes, or equivalent agent frameworks
• Strong backend engineering experience (PostgreSQL required)
• API integration experience in real production environments
• Strong proficiency in Python or TypeScript/JavaScript
• Comfortable working in CLI, Git, and deployment environments
• Ability to independently design and implement full systems end-to-end

Tech Stack

• JavaScript
• Postgres
• Python
• TypeScript