AI Operations Builder — Customer Operations

Homeone Water Heater Pte Ltd · Entry / Junior level · Full-time · SGD 4,500–7,500/month(experience-dependent) · Homeone Office, 2 Kallang Pudding Road

What this role actually is

You turn a busy, reactive customer service operation into a system that runs on rules instead of people. Not answering enquiries faster — building the machine that answers them, and running it.

Homeone installs and services water heaters, gas systems, aircon, electrical and plumbing across Singapore. A hundred-plus jobs a day, each one throwing off enquiries, scheduling, follow-ups, complaints, warranty questions.Today a person handles each of those from scratch, every time. Your job is to fix that at the root: build an AI-driven system that handles the repeatable 80% automatically and consistently, and routes the 20% that genuinely needs a human to the right place.

You start by doing the work the old way — answering calls, handling enquiries, dealing with complaints face to face. Not busywork: it's how you learn what the system needs to do, from the inside. Then you spend your shifts building the machine that makes the manual version obsolete.

Why this role exists — and why it matters for your career

Every large company already runs on this — banks, tech firms, telcos. Staff train the AI on their systems; then the AI runs those systems and the staff move up or on. This role is that shift, early, at a company small enough that you build the thing rather than inherit it.

That means you become the person who controls the AI, not the person the AI replaces. You build it, improve it, and understand the domain well enough to know when it's wrong and fix it. In two years you'll have implementation experience most people your age won't have — because you'll have built the tools, not been handed them.

What you will do

• Map the customer service flow end to end: where enquiries come from, how they're handled now, what gets missed, what takes too long.
• Sit with experienced operations staff and extract the rules they apply on instinct — the judgement that isn't written down anywhere.
• Build AI workflows that triage, prioritise and route customer cases automatically — using tools like Respond.io, ChatGPT, Claude, and whatever else earns its place.
• Watch performance daily: what the AI is handling well, what's slipping, which cases need a human.
• Report weekly on coverage, response times, and exceptions. Every week should be better than the last.
• Own the output. When the system works, that's yours. When it's wrong, you catch it and fix it.

What we are looking for

Fresh graduates and career changers with the right instinct are welcome. No customer-service background required — we can't teach instinct, but we can teach the technical side from zero. We are filtering for:

• Systematic thinking. You see a messy process and your first instinct is to understand its structure — not just to do it faster.
• Curiosity about how people make decisions. When an experienced person handles something on instinct, you want to know why they did it that way, not just what they did.
• Comfortable being wrong. You can build a first draft, be wrong, and fix it. You don't need external validation before you'll put something in front of someone — you need it to be accurate and functional.
• Fast and self-directed. Problems from today are stale by next week. The operation does not wait.

On AI use

This role is daily, heavy AI-tool use — ChatGPT, Claude, Respond.io,whatever gets the job done. We expect you to use AI to research, structure,automate, and draft. What we expect you to add is judgement: knowing that the right question is, catching when the AI output is wrong, and making the system smarter over time.

If you've never seriously used AI tooling, that's not a barrier. If you're afraid of it or dismissive of it, this is not the role for you at this stage. If you've used AI already and know its limits, we want to hear from you.

If you want to go deeper: we use Claude Code — Anthropic's command-line AI coding tool — in our operations. If you're keen to learn it, we'll teach you. It's not a requirement for the role, but it's a fast track to building automation workflows at a level most people only read about. Few companies offer this exposure to an entry-level hire.

If you have interest or skills in graphic design, video editing, or content creation — these will be put to real use. Our marketing and content team produces daily material across multiple channels, and you can contribute across operations, not just within a single job description.

What Homeone will provide

• Full training on AI tools, customer service systems, and our operating processes — you learn by doing, not by sitting in a classroom.
• Direct access to experienced operations staff who will be your domain knowledge source.
• A live operating environment where your work goes into production — not a sandbox, not a simulation.
• Permission to make mistakes. Our culture is lean: if you're not making mistakes, you're not building fast enough.
• Clear direction from the owner on what matters and why.

What will not work here

• People who need to be told exactly what to do each morning — you'll need to figure out what matters and act on it.
• People who are uncomfortable with AI tools, or believe they are a passing trend.
• People who need a large team or corporate structure around them to feel productive.
• People who treat customer service as a holding pattern until a "real" job comes along — this is the real job.

How to apply

Send your application to chpeck@homeone.com.sg with the subject line: AI Operations Builder Application.

Include the following — nothing more, nothing less:

1. A short note (200–400 words) on why this role and why now — be specific, not generic.
2. One example of a time you took a messy or unclear situation and created order from it — from work, school, personal, anything counts.
3. Your current or most recent AI tool stack — what you use and how.
4. Your LinkedIn profile or a short work history.