Operational Systems Engineer

What you would own, end to end

You take a scoped feature from a buildable spec and carry it to a production release that another person can maintain, including you six months from now. For a physiotherapy clinic, that is the patient intake flow that feeds the AI extraction, the review screen where a clinician approves or corrects what the AI pulled, and the export that goes to the practice-management system. For a notary office, it is the document queue, the interface that surfaces a cited answer from the retrieval layer, and the audit trail that logs every query. For a trades company, it is the job-request dashboard and the mobile view a field technician uses in a truck with one bar of signal.

The definition of done is a user depending on it, not a pull request merged. You own the release, the post-release monitoring, and the bugs that come back from production. That accountability is not a burden; it is what makes the work worth doing well.

The work and why it pulls

The operational systems here are the scaffolding that turns AI capability into something a business can trust. A model can extract structure from a document, but only if there is a queue feeding it, a permission layer controlling what it sees, a UI that surfaces the result, and a review path that closes the loop when the model is wrong. All of that is engineering, and all of it determines whether the clinic's staff will actually use the system or route around it.

The stakes are concrete. If the queue that feeds the AI extraction breaks, the clinic's morning intake stops. If the mobile interface is unusable on a weak connection, the technician writes it on paper and the data is lost. If the review screen is confusing, a clinician approves an incorrect record because correcting it takes too long. You are building the parts of the system where failure is felt by the people who use it, not just logged in a monitoring dashboard.

In your first months, expect to build a complete feature across web or mobile from scoped spec to production, own the data model and the API layer it sits on, and ship the review interface that a non-technical user will use to audit what the AI produces.

Who you would work with and where it leads

You work alongside the AI Engineer who is building the features your system surfaces, and a Solutions Architect who has already mapped the client's workflow and scoped the build before it reaches you. The team is small and senior. The expectation is that you can hold the full picture of a feature in your head, from the database schema to the mobile gesture, and make sound decisions at each layer without a review chain.

The role grows toward leading the full operational layer on a build: the architecture, the platform choices, the performance and accessibility bar, and the handoff to the client that means they actually own and understand what we built. If you want to be the engineer a client trusts to carry a system they depend on every day, this is the track.

You are solid in React, TypeScript, and Postgres. React Native is the mobile layer; you do not need to be a native specialist, but you need to ship a usable, performant mobile experience. You care about the release, not just the pull request, and you have a clear mental model of where a system will break before it does.

How to get in touch

This role is one we are growing toward. Email careers@entoura.studio with a note about your background and a link to something you built. Strong candidates stay on our radar as the studio grows. If the fit looks right when the role opens, you will hear from us.