Job description
This is a remote position.
AI Agent Builder (Platform Engineer)
Remote | Philippines | Full-Time | Overlap with AEST (min. 4 hours daily) | Long-Term Remote Role
If you’re a high-performing backend or platform engineer who thrives on building real-world AI agent architectures, self-hosting secure infrastructure, and working with a fast-growing AI services firm that values engineering rigor over quick demo hacks — keep reading.
This isn’t a role for a prompt engineer waiting for instructions or making simple wrapper apps.
This is a focused, outcome-driven position where you’ll own the hosting layer, build production-grade agent runtimes, and turn custom client deployments into a scalable, reusable platform.
Why This Is a Dream Job
You’ll be working directly with the founder and AI delivery leadership for Australian professional services clients who:
Champions Innovation: Builds production-grade AI agents that integrate directly into client practice tools, infrastructure, and native AI provider accounts.
Values Your Growth: High autonomy, direct impact, and no bureaucratic layers—you’ll be building core architecture alongside elite founders.
Respects Your Expertise: Focuses on inspectable, robust software engineering rather than clever, unmaintainable code.
Builds Stability: Long-term remote engagement supported by Remote Workmate, designed for permanent stability and sustainable work-life balance.
You’ll get career transformation, high-level visibility, and the freedom to build production AI platform systems — without being stuck in a basic script-writing rut.
What You’ll Be Owning This is a core platform engineering role covering agent hosting, API integrations, retrieval pipelines, and security guardrails. You aren't just making "fun chatbots"; you are building isolated, self-hosted agent platforms on client infrastructure—mastering containerization, model pinning, spend caps, and Model Context Protocol (MCP) servers to deliver real operational value.
You’ll take the lead on:
Self-Hosted Agent Platforms: Standing up and operating isolated agent runtimes (OpenClaw or SDK orchestrators) on client-owned infrastructure and AI accounts.
Read-Only Integrations: Building thin, reusable REST and MCP-style connectors for practice management systems (Xero, FYI, Microsoft 365, etc.).
Retrieval & Knowledge Systems: Indexing client document and job data into inspectable, searchable knowledge stores.
Chat & Drafting Surfaces: Integrating agents into Teams and Outlook using human-in-the-loop draft patterns.
Guardrails as Code: Enforcing strict role-based access, audit logging, spend caps, and model pinning on every deployment.
End-to-End Handover: Writing clean hosting runbooks so mid-level engineers and clients can operate the platform seamlessly.
Requirements
This Role Is a Great Fit If You:
Are an Infrastructure & Systems Architect: You have strong Linux, Docker, reverse proxy, TLS, and self-hosting experience running systems others depend on.
Are a Production API Integrator: You have 3+ years in production backend engineering (Python or Node/TypeScript) handling rate limits, auth headers, and imperfect vendor docs.
Are an AI/LLM Operator: You have built systems calling LLM APIs in loops using tools, MCP servers, or agent frameworks, and understand where agents fail in production.
Are Security-First: You instinctually manage secrets, enforce least-privilege permissions, and keep credentials out of code repositories.
Are High-Accountability: You don't need to be chased for updates. You are reliable, communicate asynchronously, and document everything clearly.
Communicate with elite fluency in written English and can maintain at least 4 hours of daily overlap with Australian Business Hours (AEST).
What Success Looks Like
After 60–90 Days:
The Platform is Fortified: One client's agent platform is hosted, hardened, monitored, and fully documented in production.
Production is Reusable: You have deployed subsequent client environments rapidly using reusable connectors and a repeatable hosting-scope framework.
Impact is Measured: Multiple client agents are live and stable, with zero data leaks, verified spend caps, and complete audit logging.
Trust is Earned: The founder relies on your engineering judgment to scope new client infrastructure within a single day.
Final Word If you’ve ever thought: "I am a world-class engineer who wants to build real, production-grade AI agent platforms—I just want a team that values security, system design, and technical depth," this is that role.
Let’s build long-term success — together.
Originally posted on Himalayas
Who can apply
Eligible countries: United States. Accepted UTC offsets: UTC-10, UTC-9, UTC-8, UTC-7, UTC-6, UTC-5, UTC+14. Review the full description for employer-specific work authorization, residency and schedule requirements.