资深软件工程师
Staff Software Engineer
我们不只是在打造更好的技术。我们在重新定义数据的流动方式，以及世界能通过数据实现什么。在Confluent，数据不会停滞。我们的平台让信息动起来，以接近实时的方式流式传输，让公司能够更快响应、更聪明地构建，并提供与周围世界一样动态的体验。
需要某种特定的人才能加入这个团队。那些会提出难题、给予诚实反馈并互相支持的人。没有自大，没有单打独斗。只有聪明、好奇的人一起朝着更大的目标前进。
一个Confluent。一个团队。一个数据流平台。
关于职位：
你将帮助构建Confluent Cloud的AI能力——这一层让客户可以直接在其实时数据上使用AI和AI代理功能。客户无需将数据转移到单独的系统中进行推理或构建代理，而是在流数据中直接完成，作为他们已使用的同一平台的一部分，该平台用于大规模移动和处理事件。
作为一名工程师，你将负责交付该产品的重要部分——不仅仅是编写代码，还要决定该功能在组成它的各个服务中应该如何工作。这里的问题很少只出现在一个地方：比如在流数据上进行推理或构建一个对实时事件做出反应的AI代理，会同时涉及多个系统——面向用户的应用程序接口、管理模型和代理生命周期的服务、调度和运行任务的控制平面，以及实际执行推理的服务层。你需要能够跨这些边界进行思考，做出合理的设计决策，并让团队内外的工程师就方法达成一致。
你将负责：
- 设计并构建运行AI和模型推理的后端服务（主要使用Go、Java和Python）。
- 从头到尾负责功能的实现——起草设计，协调团队内外的利益相关者，并推动决策达成结论。
- 在跨越多个团队的系统上做出技术决策：模型生命周期、推理路由和代理执行。
- 负责你所交付内容的质量——代码、测试覆盖率、文档、可操作性以及发布安全性。这是为实时推理提供服务的生产基础设施，因此可靠性不是事后考虑的问题。
- 通过代码审查、设计反馈，以及成为团队信任的可以处理模糊且跨领域工作的成员，让周围的工程师变得更好。
- 参与你所在团队所负责服务的值班工作。
查看英文原文
We’re not just building better tech. We’re rewriting how data moves and what the world can do with it. With Confluent, data doesn’t sit still. Our platform puts information in motion, streaming in near real-time so companies can react faster, build smarter, and deliver experiences as dynamic as the world around them.
It takes a certain kind of person to join this team. Those who ask hard questions, give honest feedback, and show up for each other. No egos, no solo acts. Just smart, curious humans pushing toward something bigger, together.
One Confluent. One Team. One Data Streaming Platform.
ABOUT THE ROLE:
You'll help build Confluent Cloud's AI capabilities — the layer that lets customers bring AI and AI agents capabilities directly to their real-time data. Instead of moving data out to a separate system to run inference or build an agent, our customers do it in place, on streaming data, as part of the same platform they already use to move and process events at scale.
As an engineer, you'll own delivery of significant pieces of this product — not just writing code, but deciding how a capability should work across the services that make it up. The interesting problems here rarely live in one place: shipping something like inference-on-streaming-data or an AI agent that reacts to live events touches several systems at once — the user-facing API, the services that manage model and agent lifecycle, the control plane that schedules and runs the work, and the serving layer that actually executes inference. You'll be expected to reason across those boundaries, make sound design calls, and get engineers inside and outside the team aligned on the approach.
WHAT YOU WILL DO:
- Design and build the backend services (primarily Go, Java, and Python) that run AI and model inference on real-time data.
- Own features end to end — drafting the design, aligning stakeholders inside and outside the team, and driving the decision to a conclusion.
- Make the technical calls on systems that span teams: model lifecycle, inference routing, and agent execution.
- Own the quality of what you ship — code, test coverage, documentation, operability, and rollout safety. This is production infrastructure serving live inference, so reliability isn't an afterthought.
- Make the engineers around you better through code review, design feedback, and being someone the team trusts with ambiguous, cross-cutting work.
- Participate in on-call for the services your team owns, and help keep the team's processes and rituals healthy.
WHAT YOU WILL BRING:
- 10+ years of significant experience designing, building, and operating distributed systems or cloud-native backend infrastructure in production
- .Strong working knowledge of Kubernetes and distributed-systems patterns (control loops, API servers, high-scale control planes), plus the fundamentals — containerization, networking, resource isolation.
- Proficiency in at least one of Go, Java, or Python, and the willingness to work across all three.
- A track record of leading cross-team technical work: turning ambiguous requirements into designs others can rally behind.
- Excellent written and verbal communication — you can write a design doc that aligns people who don't report to you.
WHAT GIVES YOU AN EDGE:
- Exposure to model serving, LLM/agent infrastructure, or streaming data systems.
- You don't need a background in ML research or model training — this role is about building and operating the platform that serves AI reliably at scale, not inventing the models.
READY TO BUILD WHAT'S NEXT? LET’S GET IN MOTION.
COME AS YOU ARE
Belonging isn’t a perk here. It’s the baseline. We work across time zones and backgrounds, knowing the best ideas come from different perspectives. And we make space for everyone to lead, grow, and challenge what’s possible.
We’re proud to be an equal opportunity workplace. Employment decisions are based on job-related criteria, without regard to race, color, religion, sex, sexual orientation, gender identity, national origin, disability, veteran status, or any other classification protected by law.
PRIVACY STATEMENT
Confluent is an IBM subsidiary which has been acquired by IBM and will be integrated into the IBM organization. By proceeding with this application, you understand that Confluent will share your personal information with other IBM affiliates involved in your recruitment process, wherever these are located. More Information on how IBM protects your personal information, including the safeguards in case of cross-border data transfer, are available here http://ibm.com/careers/us-en/privacy-policy/.