Interfere turns a product's invisible failures into shared problems the whole team can see and fix. Every app has places where users get confused, blocked, or forced to abandon a flow, but most of those moments never make it into a support ticket. We detect those failures in real-time and provide every person responsible for the fix with the context they need. When Interfere flags a broken checkout flow, the PM uses our data to prioritize the issue, the designer sees where the experience broke down, and the engineer pulls the trace underneath. We're building the operating system for product quality, so teams can move from scattered symptoms to a shared understanding of what's actually going wrong. We're a seven-person team in New York, with $5.1M raised from Y Combinator, Vercel Ventures, Hummingbird, Designer Fund, and others. Interfere is already running in production with design partners, which means the work you ship will immediately help real teams find and fix the failures costing them users today. The category is still being defined, but the product to fill this gap is inevitable, and the company that gets there first will own how the next decade of teams ship software. We're looking for the people who will move at the speed that demands.

The Role

You'll own the data backbone of Interfere. Every signal the product reasons about (events, traces, logs, runtime behavior, code, session data) flows through systems you'll build, store, and query. The product's intelligence is only as good as the data underneath it, and that data is only useful if it's accurate, fast, and affordable at scale. That's your job. You'll work close to the AI/ML and product teams, designing the pipelines, schemas, and storage everything else sits on, and making the early architectural decisions the next ten engineers will inherit. Concretely, this looks like building:

High-throughput ingestion that handles billions of product events without dropping data, slowing down, or melting the infra bill

Real-time stream processing for detection and diagnosis, where the work has to be correct and low-latency at the same time

Storage and query systems (ClickHouse; columnar, time-series, vector, whatever the problem calls for) that stay fast as customer data grows by orders of magnitude

The indexing and retrieval infrastructure that lets agents and models find the right context at the moment they need it

Schema, taxonomy, and data-quality systems that hold up as event shapes evolve and new product surfaces appear

The cost, observability, and reliability layer for our own data systems, because observability for our customers starts with observability of ourselves

What we're looking for

You've built and operated production data infrastructure at meaningful scale, with real throughput, real cost pressure, and real consequences when it breaks

You're fluent in distributed-systems tradeoffs: streaming vs batch, consistency vs latency, full fidelity vs sampling, storage vs compute. You pick the right answer for the situation rather than the one you read most recently

You take an ambiguous data or infrastructure problem, define the next useful step, and ship without waiting for a fully specified plan

You treat cost as a feature. A system that works at 1x and burns the company at 100x isn't finished

You can explain pipeline behavior, failure modes, and tradeoffs clearly enough that engineers, AI researchers, and PMs can make the right call quickly

Nice to have

Deep experience with high-throughput streaming or stream-processing systems (Kafka, Flink, Kinesis, Materialize)

Background in observability, telemetry, or session-replay data systems

Built vector or hybrid retrieval infrastructure for AI/ML use cases

Comfort across the full data lifecycle: ingestion, transformation, storage, query, retention, deletion

Fluent in Go, Rust, Python, TypeScript, or whichever tool the throughput actually demands. We hire for how you reason about data systems, not for a specific language

Strong signals

A data system you built that other engineers relied on, and that survived contact with real load