Do we need to replace everything at once?

No. The best ERP rollouts usually start with one painful workflow, one department, or one cluster of related modules. We help you sequence the rollout so the team can adopt the system without disrupting the whole business.

Can you work with the tools we already use?

Yes. We can integrate with spreadsheets, accounting tools, CRMs, forms, payment systems, messaging, and other software that still makes sense to keep. The goal is a cleaner operating layer, not change for its own sake.

How customized can the ERP be?

The system is designed around your actual workflow, rules, approvals, roles, and reporting needs. That is the point of custom ERP: the software should match the business instead of forcing the business to behave like a template.

Who owns the code, logic, and data?

You do. Your repositories, cloud accounts, and operational logic stay under your control. We build with ownership in mind so you are not trapped inside a vendor black box.

Can we add AI agents or copilots later?

Yes. In many cases, that is exactly where we want to go. We usually recommend getting the workflow, permissions, and data structure right first. Once that base is stable, AI agents become much more useful and trustworthy.

Do you support enhancements after launch?

Yes. We can continue with follow-up modules, refinements, reporting improvements, automation layers, and AI workflow additions as the business evolves.

What if we are still running most things manually?

That is common. Manual operations are often the best place to start because the pain is visible. We can map the current process, identify the strongest leverage point, and recommend the first module or workflow to build.

All roles

Engineering·Full-time·Singapore · Hybrid·Senior

Senior Full-Stack Engineer

Build the operational systems Singapore companies run their businesses on: quoting, stock, claims, payroll, the work that actually pays the bills.

About the role

You will own real systems end to end, not tickets. A dental group books its chairs on software we wrote. A main contractor raises progress claims on it. A distributor knows what is on the shelf because of it. That is the standard: software people depend on during their working day, where a bug is not an inconvenience but a stopped invoice. You will sit close to the people who use it, because we scope from watching real work rather than from a requirements document.

The mission

Ship systems that businesses trust enough to run on, and keep them fast, clear and maintainable as they grow.

What you will own

Own a client system end to end: schema, API, interface, deployment and the decisions in between.
Sit in workflow reviews with clients and turn what you observe into a scoped, fixed-price build.
Write the data model first. Most of our systems live or die on whether the schema reflects how the business actually works.
Ship weekly. Clients see a working build every week rather than a demo at the end.
Build for the person on a tablet in a warehouse or on a wet site, not just for a desk.
Own quality: tests where they earn their place, and a bar that survives the second developer.
Handle the unglamorous correctly: GST treatment, CPF logic, permit expiry, audit trails.
Review other engineers work generously and directly.
Leave the codebase easy for someone else to take over. Our clients own their code, so it must be handover-ready by default.

What we are looking for

Five or more years building production web applications that real people used daily.
Strong TypeScript, React and SQL. You are comfortable designing a schema, not just querying one.
You have owned something in production: deployed it, watched it break, and fixed it.
You can talk to a business owner without hiding behind jargon, and come back with the real requirement.
You write clearly. Commit messages, scoping notes and handover docs matter here.
Comfortable with ambiguity: a client rarely arrives with a spec.

Nice to have, not required

ERP, accounting, inventory, payroll or field-operations experience.
Singapore statutory context: GST, CPF, InvoiceNow or Peppol.
Postgres depth, or experience with serverless deployment on Vercel.
You have built something for people who are not office workers.
AI-assisted development, agent tooling or retrieval systems.

What good looks like in your first 90 days

Week 2: you have shipped a change to a live client system and watched someone use it.
Week 6: you own a module end to end and are running your own client conversations about it.
Week 12: you have taken a build from workflow review to go-live, and the client is using it daily.

Where this role goes

Two routes, both real. Deepen technically into architecture across the client estate and the AI layer we are building on top of it. Or widen into leading builds and eventually a team, owning client outcomes rather than only code. We will ask which one you want rather than assume.

Apply: Senior Full-Stack Engineer

No cover-letter theatre. Tell us about one thing you built or ran, and why it mattered.