Software Engineer, Data Products & Platform

Decide what the data means. Evaluate sources, resolve conflicts, and model entities and relationships the rest of the product can trust.

Explore and build new data products. Start from a customer problem, prototype a model, test it against real usage, and take it through to a product experience.

Build the shared foundation. Help define how data lands, gets modeled and versioned, how releases are evaluated, and how data reaches customers.

Own how the data is served. Support interactive search, enrichment lookups, large result sets, and alerts when a tracked company or person changes, with predictable performance.

Build with models, not just around them. More of this map is built by models reading the open web and agent loops that check each other's work. You'll decide when that beats a deterministic pipeline, prove it is right, and make it economical across millions of records.

Make quality a product decision. Measure coverage, correctness, freshness, and whether an answer traces back to a source. When a model makes the call, you'll build the harness that determines whether it is good enough to ship.

Customers can ask Clay new kinds of questions and act on the answers without double-checking them elsewhere.

A new data idea can become a product experience without rebuilding the foundation.

Quality is measured against real customer use, not just whether a pipeline is completed.

The team can change how a dataset is built without breaking the products on top of it.

You have a proven track record of execution. You have 8+ years of hands-on engineering experience building products customers rely on.

You love being a product manager as well as an engineer. Our product engineers drive and own products from end to end.

You've led ambiguous, cross-functional technical work from problem framing through delivery, creating clarity and scope for others along the way.

You are an empathetic communicator. You express nuanced ideas clearly at different levels of abstraction for different audiences. In disagreements, you prioritize curiosity over confrontation, making sure everyone feels heard and understood.

You love to collaborate with others to ship high quality, thoughtful features. You talk to customers, consider multiple approaches, and work with teammates to find the right solution. You care about craft, write clear and maintainable code, and create frameworks and reusable patterns that eliminate classes of problems, not just point fixes.

Having a diversity of perspectives is important to you. You believe that having people with different backgrounds and perspectives creates a better team and a more holistic product.

You’re familiar with our current tech stack or can learn unfamiliar technologies quickly. You don't need experience with every technology listed: Product stack: React, TypeScript, Python, Node.js, Aurora PostgreSQL, Redis, ECS Fargate, Lambda, OpenSearch Data platform: Python, SQL, Iceberg on S3 with Glue and Athena, SQLMesh, ClickHouse, Airflow, Terraform Observability: Datadog, CloudWatch

Product stack: React, TypeScript, Python, Node.js, Aurora PostgreSQL, Redis, ECS Fargate, Lambda, OpenSearch

Data platform: Python, SQL, Iceberg on S3 with Glue and Athena, SQLMesh, ClickHouse, Airflow, Terraform

Observability: Datadog, CloudWatch

You've built and operated production systems where data quality and reliability both matter.

You're comfortable in SQL and Python, and can reason about data models, distributed jobs, lineage, and failure recovery.

When a customer gets a wrong answer, you treat it as your bug, wherever it came from.

When datasets disagree or the right quality bar isn't obvious, you measure the problem and improve from evidence. You'd rather bring that evidence to Product or Data Science than argue positions.

Experience with entity resolution, graph data, or temporal data.

Experience putting an LLM in the critical path of a shipped product, including how you proved it was correct.

Experience with search infrastructure or data-intensive product systems.

Experience evaluating and operating on large third-party datasets.

This isn't an analytics, BI, or warehouse-enablement role, and it isn't pure infrastructure. The team owns whether each data product is correct, useful, and fast enough for customers.