Job Description

Senior Backend Engineer, Data & Integrations

Our Client uses AI to keep the power grid safe and reliable. Much of the world's power grid is decades old, and utilities must inspect it every year to catch faults before they cause outages or wildfires.

Those inspections produce millions of drone, helicopter and fixed-wing images a year, and most are still reviewed by hand. Our PowerAI platform changes that: machine-vision models find damaged insulators, conductors, hardware, poles and towers, flag vegetation encroachment, and help utility teams decide what to fix first.

The work matters in a direct way. Every fault we help a utility find sooner is a lower chance of a blackout or a fire. You'll join a small team where engineers own real systems, see their work in customers' hands quickly, and help shape how a critical industry adopts AI.

The Systems & MLOps team:

The Systems & MLOps team runs the foundation that every one of our products stand on. Our ML team builds the models and our Platform team ships PowerAI features; we make sure the infrastructure, data pipelines and model-serving layer underneath them are secure, observable and ready to scale.

A single inspection can mean hundreds of thousands of high-resolution images moving through upload, preprocessing, AI inference and into the records utilities act on. We own the services and event pipelines that carry that data, on Google Cloud, PostgreSQL, Pub/Sub and Kubernetes.

This matters because a lost or duplicated image is a structure nobody reviews. When data moves reliably and fast, customers get results sooner and can trust that every asset was inspected.

The role:

You'll own how data moves between our services and keep our data layer fast, reliable and safe as volume grows. This is an application backend role that works closely with our infrastructure engineers: they run the cloud, clusters and observability platform, and you build the services, schemas and pipelines that run on them.

Most of the hard problems sit at the boundaries between services: message contracts, at-least-once delivery, job queue tuning, and SQL against tables with millions of rows.

What you'll own:

• Data flow. The path of image and detection data between our ingestion, processing and core API services, with idempotency and clear contracts at every hop.
• PostgreSQL. Schema, queries and migrations: indexes, query plans, locking and zero-downtime changes.
• Backend features. New API endpoints and new or changed tables for entity, process and state data.
• Design. Turn loosely defined data problems into scoped designs.
• Event pipelines. Harden Pub/Sub and BullMQ pipelines for at-least-once delivery, retries and dead-letter handling.
• Customer integrations. Reliable data paths into customer systems such as IBM Maximo, Esri ArcGIS and SSO identity providers.
• Bulk data tools. Safe, repeatable tools for moving millions of rows or images in production.
• Visibility. The metrics and tracing in your services that show where data slows down or goes missing.
• Operations. On-call for the services you build, with runbooks and incident notes others can follow.

What you'll bring:

We don't expect anyone to match every line. If you meet most of the must-haves and are excited about the work, please apply. Equivalent tools count: if you've done this work with a different queue, cloud or framework, tell us about it.

Must-haves:

• 5+ years building backend or platform systems in production.
• Production TypeScript on Node.js (NestJS, Express or similar), including async error handling and connection pools.
• Deep PostgreSQL experience: reading EXPLAIN (ANALYZE, BUFFERS), index design, transactions, isolation levels and row locking.
• Zero-downtime schema changes on large tables: concurrent index builds, batched backfills and expand-and-contract migrations.
• Event-driven integration designed for at-least-once delivery: idempotency, retries, dead-letter queues and out-of-order messages.
• Clear written communication: design docs, runbooks and incident notes.

Nice-to-haves (we can teach these):

• Moving large datasets between systems without data loss or downtime.
• Job queue tuning (BullMQ or similar) and versioned API and message contracts.
• Domain-Driven Design and CQRS: aggregates, value objects, domain events, and separate command and query paths.
• Google Cloud (Cloud SQL, Pub/Sub, Cloud Storage) and Kubernetes. AWS or Azure is a plus.
• Python. Some of our services and message contracts use Python and Pydantic.
• Geospatial or image metadata: PostGIS, EXIF/XMP and GPS data. Our data is drone imagery of utility structures.

Location and how we work:

• Location. Remote (US). You work Eastern or Pacific time hours, depending on the on-call rotation.
• Work authorization. You must be authorized to work in the United States. We can't sponsor visas for this role at this time.

Our commitment to inclusion:

Our Client is an equal opportunity employer. We welcome people of every background and make hiring decisions based on skills, experience and the needs of the role, without regard to race, color, religion, sex, sexual orientation, gender identity or expression, national origin, age, disability, veteran status or any other protected characteristic.

Research shows that people from under-represented groups often apply only when they meet every requirement. If this role excites you but your experience doesn't match every line, we'd still like to hear from you. If you need an accommodation at any point in the hiring process, tell us and we'll make it work.