You'll be the technical spine of a greenfield enterprise SaaS platform - an AI-native HR system being built for a major Southeast Asian conglomerate and eventually commercialized across the region. You'll architect it, review every line that matters, and still write code every day.
This is not a \"lead\" role where you draw boxes and hand work to others. The team is intentionally lean. You own the hard problems: the multi-tenant data model, the AI orchestration layer, the SAP integration architecture, the security posture. If a Wave misses UAT, you're the one who adds capacity. That's the deal.
What makes this role interesting: The scope is genuinely complex: 12 modules, 15+ countries, real AI agents, SAP integration, 32,000-user scale target. You'll make architecture decisions that ship. The AI tooling budget is real - every engineer has Max 20x Claude Code and ChatGPT Pro. You'll move faster than any team you've been on.
What you'll actually do
* Own the overall technical architecture - hexagonal backend services in Golang, React frontend, React Native mobile, MySQL, RabbitMQ, Redis, Docker/Kubernetes
* Design the AI agent layer (Raya and sub-agents) using LLM API orchestration and RAG - no custom model training
* Define and enforce code quality gates: SonarQube, 80%+ coverage, 2 approvals per merge
* Integrate with SAP ERP, Odoo, Coupa, and SSO via OpenID - you design these adapters, not just review them
* Mentor 3 engineers without creating dependency
What we need from you
* Golang is your primary language - not a backup, not something you used once
* You've shipped a multi-tenant SaaS product at production scale
* You understand hexagonal architecture viscerally, not just theoretically
* You use Claude Code or Cursor as a daily tool - this is a hard filter, not a preference
* You can read an English-language proposal and architecture doc comfortably
Compensation: 2,200 USD/month (PJ)