Experience

Leading supplier digitization, parts R&D platform, and AI Agent development, serving R&D, test verification, and supplier teams. Driving R&D process digitalization and AI intelligence.

Enterprise Platform — R&D Process Optimization

• Built standardized R&D template system, abstracting expert knowledge into platform rules, covering 200+ part categories and 95% of common R&D scenarios, improving cross-project experience reuse efficiency.
• Redesigned test verification document lifecycle (create → approve → archive), covering core processes. Processing time reduced from 15 days to 3 days.
• Built engineering approval automation, integrating external system data. Supported 5 vehicle models and 360 engineering approvals, reducing cross-team communication costs.
• Established standardized R&D-supplier collaboration mechanism, supporting 350 suppliers online collaboration, improving data flow efficiency and compliance.

AI Agent Development

• AI Agent: Designed and launched IDP Agent — system debug, knowledge Q&A, DB query, data analysis across high-frequency scenarios. Integrated into Feishu bot for team-wide use. Data query time reduced from 20-30 minutes to ~15 seconds.
• AI Workflow: Abstracted R&D troubleshooting into standardized workflow, built AI-assisted troubleshooting capability. Product maintenance reduced from 1.5-2 days/week to 0.5-1 day/week, overall efficiency 60%+.
• Developer Productivity: Designed and shipped 106 CLI commands, packaging high-frequency capabilities (part query, test report upload) as developer tools, improving daily R&D operational efficiency.

User research, product analysis, and feature design for overseas AI browser extension.

• Led full-cycle user research: audience segmentation, online interviews, and result synthesis. Distributed 20,000 surveys, received 187 active user responses; conducted 11 deep interviews with overseas users, identified core non-paying user needs and defined homepage quick-start strategy.
• Built product experience analytics metrics and needs analysis funnel, driving optimizations in language translation, quick toolbar, and PDF reader. After launch, uninstall rate dropped from 36.47% to 18.1%, next-day retention rose from 7.25% to 9.50%.

Product manager for autonomous driving data platform Mviz and tool platform.

• Collaborated with perception algorithm R&D and simulation teams, analyzed user behavior and requirements. Enhanced Mviz custom configuration, plugin platform capabilities, and team collaboration. Delivered 15+ features, 1,200+ users issue diagnosis efficiency improved 10x (diagnosis time from hour-level to 0.4~0.6 min).
• Independently built simulation data visualization comparison feature, significantly reducing user operation difficulty. Became a commonly used feature for 500+ downstream users in simulation scenarios.
• Deeply participated in building permission management system and plugin release platform from scratch, enabling business decoupling. Proactively implemented task resource allocation mechanism, ensuring data security and privacy, with data package processing time under 10 min.