At Zendesk, we truly believe that to build great products you have to have great people. We enjoy working with other smart, focused people who care about both the products and the code they write. We value collaboration and release frequently. We like and use agile processes and believe that pragmatism always triumphs over dogmatism. We all own the product or service we work on and enjoy the impact we have improving it.
Our mission is to elevate Zendesk’s routing and presence products to their next stage — from a rules-based engine to an agentic routing engine where intelligent, adaptive decisioning is the core differentiator. Routing in a realtime, omnichannel world is a complex problem domain that needs a robust, scalable and maintainable solution. Our recent partnership with the Machine Learning team delivered Predictive Routing to GA; this role exists because the next phase of the roadmap needs that capability to live inside the team.
We’re hiring a Staff Machine Learning Engineer to be the embedded ML expert for Routing & Presence. You’ll own the ML surface of our routing products end to end — from feature engineering and model design, through experimentation, to production serving, monitoring and continuous learning. You’ll work side-by-side with backend engineers, product managers, and the central ML platform team, and you’ll set the technical direction for how ML shapes our agentic routing engine over the next few years.
This is not a pure research role. The bar is: can you take applied ML knowledge and turn it into measurable customer outcomes, reliably, at scale, inside a product engineering team?
Own the models and algorithms at the heart of Routing & Presence. You are accountable for their quality, reliability, and interoperability with the rest of the business — from Predictive Routing today through the agentic routing engine we’re building next.
Plan and scope ML work with observability and iterability designed in from the start, and with a clear view of how the resulting algorithms overlap and interoperate with adjacent systems.
Propose and evaluate alternatives. For any meaningful design decision, surface the candidate approaches, the tradeoffs, and the reasoning — don’t jump to the first plausible solution.
Design experimentation frameworks (offline evals and online A/B) tailored to routing outcomes, with statistical rigour and a clear tie-back to customer-facing metrics.
Lead innovation sessions with Product ahead of the product development cycle — help shape what we build, not just how .
Identify new data sources needed for learning and evaluation, working with Product and Data to bring them into the pipeline responsibly.
Combine multiple signals — ticket sentiment, agent skill, historical performance, realtime load — into dynamic routing decisions, with explainability and guardrails admins can trust and tune.
Shape the boundary between classical ML, LLM-driven components, and rules ; decide when each belongs where, and why.
Anticipate issues across the full lifecycle — development, testing, deployment, operations, and support — and build mitigations in, even where you aren’t the specialist.
Work across multiple teams and systems with PMs, backend engineers, the central ML team, and adjacent product teams to define problems and set the high-level direction of solutions.
Mentor scientists and engineers — raise the team’s ML literacy, support teammates through their own challenges, and make sure learning opportunities are distributed fairly.
Look ahead 18+ months. Make sure what we design and build today still serves us as the agentic routing engine matures and the customer base grows.
Build reputation as a subject-matter expert — share best practices with the wider science and engineering community at Zendesk and beyond (writeups, talks, internal forums) where it’s useful to do so.
You are an open, thoughtful and empathetic individual, a conscientious team member, and an eager learner. You have breadth of experience across multiple contexts that you draw on to avoid past mistakes, design robust systems, and pick the best-fit solution for the situation at hand. Along with these traits, you have the experience to:
Apply classical ML (probabilistic models, feature engineering, optimisation, model selection) with a clear understanding of why an approach works, not just that it does.
Work hands-on with generative AI and LLMs — fine-tuning, RAG, prompt engineering, agentic patterns — and articulate the tradeoffs against classical approaches.
Design and run experiments end to end, including power analysis, guardrail metrics, and honest treatment of the gap between offline evals and online impact.
Ship ML in production — latency and throughput constraints, drift detection, retraining triggers, versioning, rollback, and the operational realities of models that behave differently than in the notebook.
Estimate ML project effort based on prior experience, and make product tradeoffs with your eyes open — timeline, scope, quality, and cost all on the table at once.
Communicate designs and decisions clearly to engineers and scientists at different levels, such that they can transform them into tasks and execute confidently.
Motivate research outcomes in terms of wider product strategy and long-term business goals — not just the local metric that moved.
Collaborate deeply with backend engineers; read and contribute to services written in Java and Scala, even if your primary day-to-day is Python.
Explain model behaviour, uncertainty, and limitations honestly to non-technical stakeholders, and translate product problems into ML problems (and back again).
Balance focused technical delivery with being interruptible enough to help teammates through their own ML questions and problems.
Methodically identify and decompose requirements; whiteboard problems in a group setting and apply the scientific method to a pragmatic solution.
Engage regularly in self-learning — ML and statistics fundamentals, software engineering craft, and the latest relevant research.
The Zendesk system is large and ever-evolving; our team works on key services and core components shared across the company. Our work is reviewed by peers and can be deployed to production daily, if not multiple times a day.
We predominantly work with Java , Scala , Python , Kubernetes , Spinnaker , Docker , GitHub Actions , MySQL , Redis and Kafka, alongside the ML platform tooling the central team provides (model registry, feature store, serving infrastructure, experimentation platform). You’ll have freedom to pick the right tool for the ML problem while staying within the operational guardrails the organisation relies on.
Zendesk builds software for better customer relationships. It empowers organizations to improve customer engagement and better understand their customers. Zendesk products are easy to use and implement. They give organizations the flexibility to move quickly, focus on innovation, and scale with their growth.
More than 100,000 paid customer accounts in over 150 countries and territories use Zendesk products. Based in San Francisco, Zendesk has operations in the United States, Europe, Asia, Australia, and South America.
Interested in knowing what we do in the community? Check out the Zendesk Neighbor Foundation to learn more about how we engage with, and provide support to, our local communities.
Zendesk is an equal opportunity employer, and we’re proud of our ongoing efforts to foster diversity & inclusion in the workplace. Individuals seeking employment at Zendesk are considered without regard to race, color, religion, national origin, age, sex, gender, gender identity, gender expression, sexual orientation, marital status, medical condition, ancestry, physical or mental disability, military or veteran status, or any other characteristic protected by applicable law.
By submitting your application, you agree that Zendesk may collect your personal data for recruiting, global organization planning, and related purposes. Zendesk's Candidate Privacy Notice explains what personal information Zendesk may process, where Zendesk may process your personal information, its purposes for processing your personal information, and the rights you can exercise over Zendesk’s use of your personal information.
#LI-MK12
The intelligent heart of customer experience
Zendesk software was built to bring a sense of calm to the chaotic world of customer service. Today we power billions of conversations with brands you know and love.
Zendesk believes in offering our people a fulfilling and inclusive experience. Our hybrid way of working, enables us to purposefully come together in person, at one of our many Zendesk offices around the world, to connect, collaborate and learn whilst also giving our people the flexibility to work remotely for part of the week.
As part of our commitment to fairness and transparency, we inform all applicants that artificial intelligence (AI) or automated decision systems may be used to screen or evaluate applications for this position, in accordance with Company guidelines and applicable law.
Zendesk is an equal opportunity employer, and we’re proud of our ongoing efforts to foster global diversity, equity, & inclusion in the workplace. Individuals seeking employment and employees at Zendesk are considered without regard to race, color, religion, national origin, age, sex, gender, gender identity, gender expression, sexual orientation, marital status, medical condition, ancestry, disability, military or veteran status, or any other characteristic protected by applicable law. We are an AA/EEO/Veterans/Disabled employer. If you are based in the United States and would like more information about your EEO rights under the law, please click here .
Zendesk endeavors to make reasonable accommodations for applicants with disabilities and disabled veterans pursuant to applicable federal and state law. If you are an individual with a disability and require a reasonable accommodation to submit this application, complete any pre-employment testing, or otherwise participate in the employee selection process, please send an e-mail to peopleandplaces@zendesk.com with your specific accommodation request.
ZendeskRemoteYou’ll have freedom to pick the right tool for the ML problem while staying within the operational guardrails the organisation relies on.…30d