Backend Services Engineer Temporary (Full-Time, 6-Months)
KPM Analytics is a global leader in scientific instrumentation, dedicated to helping companies manage product quality and brand value across diverse sectors, including food, agriculture, industry, environment, and clinical applications. Our innovative solutions empower clients to meet critical quality parameters and enhance their operational effectiveness.
We build industrial machines that make real-time decisions on production lines around the world. Every one of those machines produces data that matters after the moment has passed: measurement records and images, configuration changes, and health signals. We are building the cloud services that receive all of it, keep it safe and separated per customer, and serve it back to the people who need it: customers, field engineers, and our own fleet teams.
Summary
This is a delivery role on the service layer: the APIs and data services that machines in the field and people in the office talk to. The engineer is being brought in to put production services on the ground in a fixed six-month window, working from written interface contracts and recorded designs.
Fleets of machines deliver data inbound over authenticated connections; customers and field engineers read it back through tenant-scoped APIs and views. The service layer sits between: receiving and storing what machines send, enforcing per-customer isolation on every read and write, and presenting stable, versioned interfaces to the consumers above it, all running on a cloud platform built alongside. This is not greenfield: the machine-facing interfaces exist as written, versioned contracts with working implementations already in production use.
On-site - Westborough, MA. Full-time, hourly, temporary, 6-Month position.
Duties and Responsibilities
- HTTP APIs and supporting services designed, implemented, tested, and deployed against written interface contracts, with idempotent handling of retries and replays as a design rule, not an afterthought.
- Relational schemas, migrations that roll forward safely, and query paths that hold up as data grows; per-customer scoping enforced in the data layer itself and proven by automated tests.
- OIDC-based identity for people, certificate- or token-based identity for machines, and role-based access enforced server-side.
- The views and export paths that let customer tools consume data without touching internal tables, stable across internal schema change.
- The ingest, retention, and read paths for the health and status signals machines already report, scoped and access-controlled as an internal-facing surface distinct from the customer-facing data services, so a stalled machine is seen internally before the customer calls.
- Each service arrives as a container image with its pipeline, configuration, and forward-rolling migrations described in code and gated in CI; nothing shipped is installed or configured by hand.
- Unit and integration tests wired into CI, so a green check is a deployable claim.
- Interface docs, decision records, and operational notes written as work proceeds, reviewed like code. The engagement is not complete until someone else can run and extend what was built.
Qualifications Required
- 2+ years of professional experience building and operating backend services that ran in production for real users.
- HTTP APIs designed and operated in production, including ownership through design, implementation, versioning, and the support of real consumers, able to explain failure modes such as retries, timeouts, idempotency, and backward compatibility.
- Relational databases in production, schema design, migrations, indexing and query performance, and the discipline of least-privilege database accounts.
- Authentication and authorization implemented in earnest, OIDC/OAuth2 or JWT flows, TLS, secrets handling, and server-side access control written, not just configured.
- Proficient programming in at least one backend language, with production code that can be walked through and defended line by line.
- Linux, Git, and CI/CD as daily working tools.
- AI-assisted development ("vibe coding") as a working skill, direct experience directing an AI coding agent through real, multi-step work, with critical review, testing, and correction of its output.
- Independent delivery, able to take a written contract and a deadline, decompose the work, surface risks and blockers early, and hand over finished, documented pieces on a cadence without day-to-day direction.
- Eligible to work on-site in Westborough, MA as a contractor for the full six-month term.
Language Skills
Clear written communication is required, along with professionalism, to have work reviewed critically and to review others’ work the same way. All documentation, decision records, and interface contracts are written and read in English.
Education
No specific degree required.
Traits/Skills preferred
- Streaming or event-driven systems in production (Kafka or equivalent).
- Device, IoT, or fleet-connected data, high-volume ingest from many remote producers.
- Multi-tenant data isolation done in earnest.
- Python services in production.
- Containerized deployment and the operational habits that go with it.
- mTLS or certificate lifecycle work.
- An observability stack instrumented personally.
- Media or binary-object handling at volume.
Work Environment
- Work proceeds from written interface contracts and recorded decisions, read before changing things, and written the same way.
- Everything moves through pull requests and review, including documentation. No direct-to-production changes.
- Weekly delivery checkpoints against an agreed plan; risks and slips are raised when they appear, not at the deadline.
- Handover is a deliverable from week one, not an activity reserved for the final week.
- The service layer is delivered against a companion platform role’s environments and shared infrastructure; the seam between the two is a written deployment contract, kept current like any other interface.
Location
On-site - Westborough, MA. Full-time, hourly, temporary, 6-Month position.
