WHO WE ARE
Optiver is a global market maker founded in Amsterdam, with offices in London, Chicago, Austin, New York, Sydney, Shanghai, Hong Kong, Singapore, Taipei and Mumbai. Established in 1986, today we are a leading liquidity provider, with close to 2,000 employees in offices around the world, united in our commitment to improve the market through competitive pricing, execution and risk management. By providing liquidity on multiple exchanges across the world in various financial instruments we participate in the safeguarding of healthy and efficient markets. We provide liquidity to financial markets using our own capital, at our own risk, trading a wide range of products: listed derivatives, cash equities, ETFs, bonds and foreign currencies.
Since its establishment in 2012, our Shanghai office is a rapidly growing participant in the Chinese markets, trading exchange-listed futures, options and equities in China mainland. Our vision is to become the trusted partner in the development of Chinese financial markets. With the culture of a start-up but the backing of a 35+ year-old trading firm, the Optiver Shanghai office is truly unique. Everyone who joins us will help shape the future of our company and its global impact. Get ready: we are only just beginning.
ABOUT THE TEAM
Optiver is standing up a new AI Engineering team in Shanghai to own model strategy for the firm globally: which models we run, where they run, what they cost, and the proof that every model swap is safe. The team builds and operates a diversified model substrate spanning frontier APIs (Anthropic via Bedrock, Vertex, and first-party) and open-weight models (GLM, DeepSeek, Qwen/Kimi) hosted on third-party cloud and on-prem, alongside the evaluation machinery that makes model selection and routing decisions defensible. You'll grow the team from its first engineers to 10–15 people, organised into two co-located, interchangeable functions: model serving and evaluation. You'll set the team's ceiling.
WHAT YOU'LL DO
- Build the team from 2–3 engineers to 10–15 across two pods (model substrate and evaluation), setting the hiring bar, running the interview slate with Shanghai recruitment, and deciding the team's shape as it scales.
- Own the diversified model substrate: onboard and operate open-weight models on third-party cloud and on-prem inference, and keep the full provider matrix live (Anthropic via Bedrock/Vertex/first-party; GLM, DeepSeek, and Kimi via Alicloud; self-hosted open weights).
- Reduce model supply risk from export restrictions by making the substrate a credible hedge, with regionally-accessible options proven equivalent for the tasks that matter.
- Benchmark models per business function and give users a clear answer to "which model for this task," including guidance around model upgrades.
- Build the evaluation machinery that makes swaps provable: shared graders, calibrated judges, a harvest-and-curate loop for eval data, and a certification gate that replays each agent's suite against a candidate model (passes auto-upgrade, regressions get pinned).
- Own model routing and inference cost optimisation, closing the loop from spend visibility (which this office already holds) to routing decisions proven safe by evals.
- Work across AI Risk, Procurement, Developer Experience, and China office leadership, and partner with the global AI Engineering leads on substrate buildout from day one.
WE'RE LOOKING FOR CANDIDATES:
- With deep production experience serving and tuning large models: inference optimisation, GPU capacity planning, and hosting open-weight models (GLM, DeepSeek, Qwen, or similar) on cloud and on-prem.
- Who have built or run model evaluation systems at scale: benchmark design, LLM-as-judge calibration, and regression gating for model upgrades.
- Who have grown an engineering team from a handful of people to 10 or more, including setting and holding the hiring bar.
- Who reason about models as cost/capability trade-offs and can defend a routing decision with eval data rather than intuition.
- With fluent Mandarin and English, comfortable operating with senior stakeholders across US, EU, and APAC time zones.
- Senior enough to be the first hire and set the bar: this person owns a live deliverable (the substrate buildout) from day one.
WHAT YOU'LL GET
- A performance-based bonus structure unmatched anywhere in the industry. We combine our profits across desks, teams and offices into a global profit pool, fostering a truly collaborative environment.
- The chance to work alongside diverse and intelligent peers in a rewarding environment.
- Training, mentorship and personal development opportunities.
- Daily breakfast, lunch and snacks.
- Gym membership, sports and leisure activities, plus weekly in-house chair massages.
- Regular social events, clubs and Friday afternoon drinks.
DIVERSITY STATEMENT
Optiver is committed to diversity and inclusion. We encourage applications from candidates of all backgrounds, and welcome requests for reasonable adjustments during the process.
PRIVACY DISCLAIMER
Optiver 重视个人信息的保护。请您在提供个人信息给我们之前，认真阅读Optiver China Privacy Notice, 了解我们如何收集及处理您的个人信息。
Personal information protection is of utmost importance to Optiver. Before you provide any personal information to us, we strongly urge you to read our Privacy Policy to acknowledge how we collect and process your personal information.
Skills Required
- Deep production experience serving and tuning large models including inference optimization and GPU capacity planning
- Experience hosting open-weight models (GLM, DeepSeek, Qwen or similar) on cloud and on-prem
- Built or run model evaluation systems at scale: benchmark design, LLM-as-judge calibration, regression gating
- Proven experience growing an engineering team from a handful to 10+ engineers, setting and holding hiring bar
- Experience designing model routing and inference cost optimization tied to evaluation results
- Fluent Mandarin and English, comfortable operating with senior stakeholders across US, EU, and APAC time zones
- Senior leadership capable of being the first hire and owning live deliverables from day one
- Experience working with cloud providers and vendor integrations (e.g., Anthropic via Bedrock/Vertex, Alicloud)
Optiver Compensation & Benefits Highlights
The following summarizes recurring compensation and benefits themes identified from responses generated by popular LLMs to common candidate questions about Optiver and has not been reviewed or approved by Optiver.
-
Career-Linked Recognition & Rewards — Pay is considered highly competitive in trading, quantitative, and engineering paths, with performance-linked bonuses materially elevating total earnings for strong contributors. High performers are described as seeing outsized upside when firm and team results are strong.
-
Healthcare Strength — Health coverage is depicted as comprehensive, including medical, dental, vision, disability and life insurance, alongside HSA/FSA options and mental-health support. Some locations note unlimited therapy access through a partner platform.
-
Leave & Time Off Breadth — U.S. roles are often described as offering 25 days of paid vacation plus market holidays, alongside generous parental leave. This breadth of time off is highlighted as helping to offset a high-intensity environment.
Optiver Insights
What We Do
Optiver’s story began over 30 years ago, when we started business as a single trader on the floor of Amsterdam’s options exchange. Today, we are at the forefront of trading and technology as a leading global electronic market maker, focused on pricing, execution and risk management.
Why Work With Us
People at Optiver love challenges, welcome collaboration, and strive to be better tomorrow than they are today. Improving the market is an extraordinary challenge that requires a carefully crafted approach. Optiver provides a collaborative working environment to tackle these challenges. In fact, it is this way of working that sets us apart.
Gallery
