Senior Software Engineer, Data Product

Shippo

30d•Remote

About The Position

We are seeking a Senior Software Engineer to join our Data Product team. In this role, you will own the backend services responsible for delivering Estimated Delivery Date (EDD) predictions to merchants and internal consumers. You will build high-throughput, low-latency Python services, lead API design and technical reviews for various data product surfaces including rules automation, ML-based recommendations, analytics, and configuration systems. A key part of your role will be ensuring the reliability and observability of these services through instrumentation, alerting, runbooks, and incident response. You will partner closely with data science to productionize model outputs, manage the API layer, serving infrastructure, and operational reliability of ML-powered features. Additionally, you will build and maintain feature pipelines, contribute to MLOps foundations, and instrument systems for comprehensive observability. You will also play a crucial role in evaluating frameworks, tooling, and architectural patterns for ML serving, set the technical direction for backend and ML systems, lead design and code reviews, mentor other engineers, and apply AI tooling to your own workflow.

Requirements

7+ years building production backend systems, with a meaningful chunk of that time collaborating with data and/or ML teams. You've been the engineer responsible when a model in production behaves badly at 2am.
Demonstrates ownership over large-scale projects by driving design decisions, setting scope, delegating work appropriately, and managing stakeholder expectations through execution.
Deep Python backend skills with FastAPI (or an equivalent async framework), strong PostgreSQL fundamentals (schema design, query optimization, migrations), and hands-on experience with event-driven systems like Kafka.
Track record of owning distributed systems through their full lifecycle: design, launch, monitoring, and iteration.
You know how to ship changes to production safely — canary, shadow, A/B, versioning, rollback — and can judge when each is warranted versus overkill, including for ML-backed systems.
You can instrument production systems for the signals that matter (latency, throughput, error rates), and are comfortable extending that to ML-specific signals like drift and prediction quality. You can explain to a non-ML audience what's actually wrong when one of them moves.
You write high-quality, maintainable code, own problems end-to-end from design through long-tail production behavior, and hold that standard in design and code reviews.
You communicate trade-offs clearly — including unpopular ones like "we shouldn't ship this yet" or "the bottleneck isn't the model."
You partner well with Data Science. You don't see ML as DS's job and operations as yours; you see the whole system as the team's job.

Nice To Haves

Direct experience with delivery-date prediction, ETA, or other time-series prediction systems in e-commerce, logistics, or transportation.
Domain experience in shipping, logistics, carrier APIs, or rate selection.
Experience working in or alongside data science / ML teams — you've shipped or operated features and APIs that depended on ML models. You understand the gap between a notebook and a reliable inference endpoint.
Experience contributing to ML platform components (feature stores, model registries, serving infra) from the user side — you've made an ML platform better by being a demanding user of it.
Experience with feature stores and online/offline feature consistency.
Hands-on experience with LLM-based features, retrieval systems, or agent workflow infrastructure.
Prior experience operating in a senior engineering capacity, or stepping into informal technical leadership on a team.

Responsibilities

Own the backend services that deliver EDD predictions to merchants and internal consumers — APIs, caching, contracts, and reliability under production load.
Build Python services suited to high-throughput, low-latency workload.
Lead API design, service decomposition, and cross-team technical reviews for data product surfaces spanning rules automation, ML-based recommendations, analytics, and configuration systems.
Own reliability and observability across the services you build—instrumentation, alerting, runbooks, and incident response.
Partner with data science to bring model outputs into production—owning the API layer, serving infrastructure, and operational reliability of ML-powered features.
Build and maintain feature pipelines that bridge offline training and online inference, with an emphasis on consistency and data quality.
Contribute to MLOps foundations for the team: model deployment patterns, versioning, rollback procedures, and experiment tracking integrations.
Instrument systems for observability—latency, throughput, drift signals, and prediction quality—so issues surface before they reach merchants.
Be a voice in evaluating frameworks, tooling, and architectural patterns for ML serving and make pragmatic recommendations grounded in production experience.
Set the technical direction for backend and ML systems on the Data Products team—proposing and driving architectural decisions that balance velocity with long-term maintainability.
Lead design reviews, raise the bar in code reviews, and establish engineering practices the team can follow.
Mentor other engineers on backend and systems engineering.
Apply AI tooling to your own workflow and share learnings with the team.