Own the whole problem
Stay with the work from problem definition through deployment, rather than delivering one isolated module.
01 / The mission
Build AI systems teams can rely on
Turn product requirements into agents, data flows, and systems integrations. Build evaluation, monitoring, and feedback mechanisms, then improve reliability, usability, and engineering quality based on actual use.
02 / What you'll own
Frame the problem
- Join project kickoffs to understand current systems, data, operating constraints, and execution.
- Work with Product FDE to turn a high-value problem into a clear, buildable system boundary.
Build and integrate
- Design and build agents and data flows that connect research, judgment, and action.
- Connect agents and data flows to the tools and business systems the team already uses.
Evaluate and reuse
- Establish evaluation, monitoring, and feedback loops to improve reliability, usability, and engineering quality.
- Turn validated engineering and execution patterns into reusable capabilities and stay accountable for real operation.
03 / What we look for
- Strong engineering skills and the ability to make progress when requirements are still taking shape.
- A willingness to understand customer, business, and life-science problems rather than receive technical tickets only.
- Attention to how people use the system and whether it helps them do their work.
- The ability to use and keep learning AI tools.
- The ability to work in Chinese or English and collaborate with teams and customers in a bilingual environment.
- A willingness to work on site with customers at kickoff and key project stages.