AI-Integrated IT & Ops Engineer (Agentic Infrastructure)

SENSLAB HK LIMITED; Hong Kong; On-site; HKD 420,000 - 600,000; Full time

Job summary

SENSLAB HK LIMITED is seeking an AI-Integrated IT & Ops Engineer to lead the integration and management of internal AI agents within our infrastructure. The role focuses on building scalable automation, teaching AI agents to operate AWS, Docker, and Linux environments, and ensuring secure, delegated workflows across corporate IT tasks.

You will script, document, and test agent capabilities, while maintaining robust system configurations and zero-trust boundaries.

Qualifications

• 3+ years of experience in DevOps, IT Operations, Sysadmin, or SRE.
• Strong hands-on experience with AWS, Linux (Ubuntu/Debian), Docker, and Git.
• Experience with internal corporate software administration (Feishu/Lark, Google Workspace, Slack).
• Familiarity with modern AI tools, APIs (OpenAI, Anthropic, Gemini), and multi-agent systems.

Responsibilities

• Agent Enablement & Skill Building: Write Python/Bash scripts and clear Markdown Skills to teach AI agents to manage IT environment, provision accounts, and orchestrate cloud resources.
• Infrastructure Management: Maintain AWS EC2 instances, Docker sandbox environments, Tailscale networks, and Linux configurations for secure agent operation.
• Corporate IT & Lark/Feishu Admin: Oversee IT operations by delegating tasks to AI assistants via custom API integrations.
• Escalation & Troubleshooting: Debug networking, API timeout, and Linux permission issues when AI agents fail.
• Security & Zero-Trust: Enforce sandbox boundaries, API whitelists, and process isolation to prevent host compromises.

Skills

Python

Bash

Linux

AWS

Docker

Git

LLMs

Tools

Feishu/Lark

Google Workspace

Slack

Job description

AI-Integrated IT & Ops Engineer (Agentic Infrastructure)

We are looking for an IT & Operations Engineer with a completely new mindset. In this role, you will not just manage servers, employee accounts, or AWS instances—you will manage our internal AI agents (like our dedicated corporate assistant, "Ariel").

Your primary philosophy must be: Delegate, Teach, then Verify. When a new infrastructure problem arises or a new ops process is needed, you won't just SSH into the server and fix it manually. You will write a script, wrap it in a Markdown-based "Skill," teach the AI agent how and when to use it, delegate the execution to the agent, and verify the logs. You are the Commander of our digital workforce.

Key Responsibilities

• Agent Enablement & Skill Building: Write Python/Bash scripts and clear, context-dense Markdown documentation (Skills) to teach our local AI agents how to manage our IT environment, provision accounts, and orchestrate cloud resources.
• Infrastructure Management: Maintain our AWS EC2 instances, Docker sandbox environments, Tailscale private networks, and Linux system configurations (Systemd, permissions, namespaces) to ensure a secure operating environment for our agents.
• Corporate IT & Lark/Feishu Admin: Oversee corporate IT operations, but primarily by delegating tasks to our AI assistants through custom API integrations and webhook events.
• Escalation & Troubleshooting: Jump into the terminal when the AI agents fail. Debug complex networking, API timeout, and Linux permission boundary issues (the "90% Linux sysadmin, 10% prompt engineering" reality of production AI).
• Security & Zero-Trust: Enforce strict sandbox boundaries, API whitelists, and process isolation (e.g., Tmux execution rules) to ensure autonomous agents cannot compromise the host system.

Who You Are

• The Scripting Hacker: If you do something twice, you script it. You are highly proficient in Python and Bash.
• The Linux Wizard: You understand the Linux kernel primitives, Docker socket permissions, SSH tunneling, and Systemd cgroups. You know why a permission denied error happens when a background daemon touches a socket.
• The AI Whisperer: You have a strong intuition for Large Language Models (LLMs). You understand the difference between a system prompt and a user prompt, how context windows work, and why rigid "ALWAYS DO THIS" rules are inferior to explaining the "Why" (Theory of Mind) to an AI.
• The Technical Writer: You write incredibly clear, concise documentation. You believe "the context window is a public good" and know how to strip out fluff so both humans and AI can parse your instructions instantly.

Requirements

• 3+ years of experience in DevOps, IT Operations, Sysadmin, or Site Reliability Engineering.
• Strong hands‑on experience with AWS, Linux (Ubuntu/Debian), Docker, and Git.
• Experience with internal corporate software administration (Feishu/Lark, Google Workspace, Slack, etc.).
• Familiarity with modern AI tools, APIs (OpenAI, Anthropic, Gemini), and a deep curiosity for multi‑agent systems and Agentic OS architectures.