> Markdown version of [/jobs/ext/3912767-software-engineer](https://www.wearedevelopers.com/jobs/ext/3912767-software-engineer). Every page supports `.md` or `Accept: text/markdown`. Links point to the HTML versions so they work for humans too. Agent guide: [/agents.md](https://www.wearedevelopers.com/agents.md).

---

# Software Engineer

- **Company:** Turquoise Health
- **Location:** San Diego, CA, United States (Remote available)
- **Experience:** Expert
- **Contract:** Permanent contract
- **Skills:** Application Programming Interfaces (APIs), Big Data, Cloud Computing, Databases, Data Architecture, Data Stores, Data Systems, Data Warehousing, Python (Programming Language), Load Testing, Operational Databases, Sql Optimization, Concurrency, Backend, Front End Software Development, Vertica, Pagination
- **Published:** October 26, 2026
- **Apply:** https://startup.jobs/senior-software-engineer-performance-turquoise-8800495

## About the Role

* 5+ years building and optimizing production data systems, data-intensive applications, or analytical products
 * Advanced SQL and strong Python (or another general-purpose language)
 * Hands-on experience optimizing analytical workloads in column-oriented databases or data warehouses
 * A hypothesis-driven approach to performance work: you use query plans, profiling, benchmarks, and production telemetry to define the problem, set a baseline, test alternatives, and quantify the result in production
 * The ability to reason across system boundaries (databases, APIs, backend services, client request patterns, product interactions) and the product judgment to know when changing the experience beats changing the data system

Bonus points

 * Experience with ClickHouse
 * Experience contributing to data-heavy search, filtering, analytics, or reporting products
 * Experience with healthcare specific data types: claims, contract rates, revenue cycle, medical billing, or healthcare pricing experience

## Description

For our Analyze and Search products, speed is part of the product. Our customers filter, compare, and benchmark across large healthcare pricing datasets, and a slow query changes how much they trust and use what we've built. You'll make that experience fast, scalable, and cost-efficient as datasets grow and more products build on the same platform.

The challenge is that performance depends on the combination of filter, dataset, and workflow, not on any single query. A dataset update can shift latency before anyone notices, and the best fix is sometimes a better data model and sometimes a better product behavior. You'll investigate, benchmark, prototype, and ship across the path from ClickHouse to the API to the screen, working directly with Product, frontend, and backend engineers.

This is a data and product performance role, not a load-testing, QA, or low-level systems role. If you like turning "this feels slow" into a baseline, a hypothesis, and a number that moves in production, you'll enjoy it here.

Responsibilities

Diagnose and fix

 * Own performance investigations across the product and data stack, taking a slow query or poor user experience to a measured root cause and a production fix. Example: a critical filter and data-source combination is now x% faster in production.
 * Track latency, throughput, concurrency, and infrastructure cost, with a focus on datastore efficiency (we run on ClickHouse Cloud). Example: a product workflow now uses z% fewer ClickHouse resources without compromising the user experience.

Anticipate

 * Build representative benchmarks for the filter, dataset, and workflow combinations that matter most.
 * Evaluate the performance impact of upstream dataset changes before they reach production. Example: an incoming change is expected to increase latency by y%, and we have a mitigation plan before launch.
 * Assess future product requests and recommend technical or product approaches that balance functionality, responsiveness, scalability, and cost.

Design and ship

 * Design and evaluate data architecture changes, such as restructuring tables, precomputing results, and changing indexes.
 * Partner with Product, frontend, and backend engineers to adapt product behavior to the data, using techniques like request batching, debouncing, caching, and pagination.
 * Prototype competing solutions, communicate their tradeoffs clearly, and validate the chosen approach in production.

Build for the long term

 * Improve performance instrumentation, monitoring, and regression detection across data-powered product experiences.
 * Establish reusable patterns as the Data Product platform becomes a supporting layer for more products and teams.