Overview
Own ClickHouse's official Go clients (clickhouse-go and ch-go) and the collectors, pipelines, agents and tools built on them, deciding what the clients should do by working with the developers who use them.
What you'll do
- Know who uses the clients and what they're trying to do; define what good looks like (time to first query, error clarity, throughput, adoption by key integrations), measure it, and let that drive the roadmap.
- Be the voice of ClickHouse for Go developers: speak at GopherCon and Go meetups, write posts, tutorials and examples (SEO and GEO), and answer community questions.
- Own the Go clients and their ecosystem (clickhouse-go, ch-go, database/sql, releases and support), evolving them without breaking dependents.
- Keep the clients the fastest Go option for the fastest analytical database and ensure the client is never the bottleneck: profile, change, measure again, catch regressions in CI, and publish the numbers.
- Partner with projects built on the clients, from the OpenTelemetry Collector to the ClickHouse Grafana datasource and ClickPipes, and catch their regressions before they ship.
- Review and co-maintain other clients and take part in cross-client decisions on types, errors and protocol features.
- Bring written proposals about where to invest next and set direction rather than wait for it.
- Make the clients enterprise-ready: TLS and mTLS, cloud authentication, safe credential handling, vulnerability response, a small dependency footprint, and the questions large Go shops ask before adopting.
- Shape the clients for agents as users: an API, docs and error messages an AI coding agent can use correctly on the first try.
- Extend the AI development platform with workflows for maintaining the clients: consistency checks across languages, issue triage, release checks.
What you'll need
- 5+ years in software, much of it on data-intensive systems: drivers, SDKs, ingestion, streaming, observability or analytics.
- Deep Go expertise: the runtime and garbage collector, escape analysis, the memory model, pprof and the execution tracer.
- Reason about throughput and latency down to allocations and bytes on the wire, and prove it with tests and benchmarks.
- Strong Go concurrency instincts: context, goroutines and channels, bounded pools, backpressure, clean shutdown, and the leaks and races that come with each.
- Beyond Go, productive in at least one more of the client languages (Python, JavaScript, Java, Rust, C++, C#).
- Built or substantially extended a Go module other engineers depend on; understand semantic import versioning and treat breaking changes accordingly.
- Strong grasp of database fundamentals: SQL, query execution and analytical workloads, with judgment about when to use database/sql and when to drop down to the protocol.
- Enthusiastic about working in public: given talks or written technical content, or keen to.
- Thoughtful use of AI development tools, with the judgment to review what comes out.
- Clear writing; most coordination happens in design docs, PRs and RFCs.
Nice to have
- ClickHouse, especially its type system and native format, or another columnar or OLAP engine.
- Contributions to Go data or observability infrastructure: OpenTelemetry, Prometheus, Grafana, Kafka clients, database drivers.
- Building or maintaining a database/sql driver.
- Apache Arrow or ADBC in Go.
- Experience designing SDKs, docs or error messages with AI coding agents as the consumer.
- Protocol or wire-format work on a database driver or network client.
Details
- Location: Canada (remote-friendly; the team is remote across North America and EMEA).
Read the full description and apply on the company’s own careers page.