Senior Full Stack Engineer, Account Linking

Senior-level engineers with 7+ years of professional experience in full-stack development, particularly with Node.js, React or React Native, GraphQL, and PostgreSQL or similar databases, are suitable for this role.

They should be skilled in building scalable integrations, asynchronous workflows, data observability, reliability, security, and data accuracy, with experience in tools such as DataDog, BigQuery, SQS, and Temporal being valuable.

The ideal candidate is familiar with owning complex features end to end, supporting high-scale financial data ingestion, managing migrations and duplicate data, troubleshooting performance and user experience issues, and contributing to technical design, RFCs, code reviews, and cross-functional collaboration.

How would you design a reliable financial-account integration service using Node.js, PostgreSQL, SQS, and Temporal?

I would separate ingestion, normalization, persistence, and analysis into independently observable workflows. Temporal would manage durable, retryable orchestration, while SQS could buffer high-volume events; PostgreSQL would enforce transactional consistency and idempotency through unique keys and state transitions.

How would you make account and transaction ingestion idempotent?

I would derive stable provider-specific identifiers and enforce unique constraints on account and transaction records. Each workflow step would also use idempotency keys and safely handle retries, duplicate messages, and partially completed operations.

How would you design retries and failure handling for unreliable financial-data providers?

I would distinguish transient errors, authentication failures, rate limits, and permanent data errors. Transient failures would use exponential backoff with jitter and bounded retries, while persistent failures would be surfaced through explicit workflow states, alerts, and user-actionable messaging.

How would you use Temporal in an account-linking workflow?

I would model linking, credential exchange, account discovery, historical sync, and incremental refresh as durable workflow steps. Activities would be short, idempotent, and independently retryable, with timeouts, compensation logic, and signals for provider reauthentication or user intervention.

How would you detect and remediate duplicate financial accounts?

I would combine deterministic identifiers with heuristics based on provider data, account type, masked identifiers, institution, and historical relationships. Potential matches should be scored and reviewed through safe merge or remediation workflows that preserve provenance and avoid deleting user-visible history incorrectly.

How would you monitor data quality for ingested financial information?

I would define metrics such as ingestion completeness, freshness, balance reconciliation, transaction duplication, schema violations, and unexpected volume changes. These signals could be aggregated in BigQuery, monitored in Datadog, and evaluated with heuristic thresholds that account for provider-specific behavior.

How would you design a GraphQL API for linked accounts and financial data?

I would expose a stable, domain-oriented schema with clear authorization boundaries and pagination for large collections. Resolvers should avoid N+1 queries through batching and caching, enforce field-level access controls, and return explicit synchronization status rather than hiding eventual consistency.

How would you optimize PostgreSQL for high-volume transaction ingestion and querying?

I would use appropriate composite indexes, bulk inserts, carefully chosen upserts, and partitioning when data volume or retention requirements justify it. Query plans, lock contention, connection pool usage, and write amplification should be monitored before and after changes.

How would you handle sensitive financial data securely across the React Native and Node.js stack?

I would minimize collection, encrypt data in transit and at rest, avoid logging credentials or sensitive payloads, and enforce least-privilege access. Mobile tokens should use secure platform storage, while backend access should include strong authorization, audit logging, retention controls, and careful secret management.

How would you improve the user experience when an account connection or synchronization fails?

I would expose clear synchronization states, distinguish recoverable issues from required reauthentication, and provide actionable retry flows. The backend should maintain durable progress and correlation identifiers so users can safely retry without creating duplicate accounts or data.

Rocket Money is hiring a Senior Full Stack Engineer to help millions of members connect, process, and understand their financial data. You’ll own high-scale integrations and end-to-end features across React Native, Node.js, GraphQL, PostgreSQL, DataDog, BigQuery, SQS, and Temporal, with a focus on reliability, data accuracy, and privacy. This is a hybrid role based in San Francisco, Washington, D.C., or New York City, offering $150,000–$185,000 annually plus bonus and benefits. Join a mission-driven team shaping innovative financial products while enjoying strong benefits, learning opportunities, stock purchasing, and unlimited PTO.