[Skip to main content](#main-content)

[ OpenNash ](https://opennash.com/) / [Jobs](https://jobs.opennash.com/) [ 01 Search jobs](https://jobs.opennash.com/search/)[ 02 Companies](https://jobs.opennash.com/explore/)[ 03 Hiring report](https://jobs.opennash.com/reports/hiring/)[ 04 AI assistants](https://jobs.opennash.com/connect/)[ 05 How it works](https://jobs.opennash.com/methodology/)

Tencent Holdings · Shenzhen

# ima-AI评测产品经理（自动化评测与工具建设）（北京/深圳）

Shenzhen 

Last seen on Tencent Holdings’s official careers site on November 2, 2026; it may have closed since.
[Apply on Tencent Holdings’s careers site →](https://careers.tencent.com/en-us/jobdesc.html?postId=2115739280528818176)

[OpenNash Jobs](https://jobs.opennash.com/) / [Tencent Holdings](https://jobs.opennash.com/companies/tencent-holdings/) / ima-AI评测产品经理（自动化评测与工具建设）（北京/深圳）

## Job description

### Responsibilities
- 围绕 ima 知识问答、检索及 Agent 等场景，建设业务评测体系，明确质量标准与评价指标；
- 建设和维护业务 Benchmark，完善样本、评分规则及更新机制，结合人工与模型辅助评测，保障评测结果可靠；
- 开发自动化评测工具，支持批量测试、自动评分、结果对比及回归验证，提升评测效率与可复现性；
- 持续追踪不同模型及版本的效果、稳定性、时延与成本，为模型选型和产品迭代提供依据；
- 建设满意度评价与反馈机制，关联离线评测和真实用户体验，协同产品、算法、研发定位问题并验证改进效果。 Requirements
- 有 AI 产品、大模型应用评测或算法策略相关经验，能够将业务问题转化为可执行的评测方案；
- 理解大模型、RAG 及 Agent 的基本机制，熟悉评测集构建、评分标准及人工／自动评测方法；
- 熟练使用AI工具，能够调用模型 API，独立搭建、调试自动化评测脚本或工具；
- 具备良好的数据分析与实验设计能力，能识别评测偏差，判断效果变化并形成清晰结论；
- 主动探索、注重用户价值，能通过动手验证和事实反馈修正判断，推动问题解决。

## At a glance
- Posted by the employer: Not stated
- First seen by OpenNash: Nov 2, 2026
- Last checked: Nov 2, 2026
- Listed for: Under a day
- Location: Shenzhen[Apply on Tencent Holdings’s careers site →](https://careers.tencent.com/en-us/jobdesc.html?postId=2115739280528818176)

## More jobs at Tencent Holdings

### [ima-PC客户端开发工程师](https://jobs.opennash.com/companies/tencent-holdings/jobs/5e1cdaf7e24d0ba35583/)

Shenzhen

### [ima-数据分析师](https://jobs.opennash.com/companies/tencent-holdings/jobs/0b8686e463b079c343a0/)

Shenzhen

### [ima-研发项目经理](https://jobs.opennash.com/companies/tencent-holdings/jobs/f617b184c4807eb8a028/)

Shenzhen

### [ima-高级产品经理（基础框架）](https://jobs.opennash.com/companies/tencent-holdings/jobs/4f0b71b71c40833105e0/)

Shenzhen

### [ima-高级产品经理（开放能力）](https://jobs.opennash.com/companies/tencent-holdings/jobs/7dcb7ea1d61ace9f7d0e/)

Shenzhen

## Similar jobs at other companies

### [AI评测产品经理](https://jobs.opennash.com/companies/meituan/jobs/b0fa4964dde974dc13a3/)

Meituan · 北京市

### [Dual Student IMA Asteasu](https://jobs.opennash.com/companies/siemens-energy/jobs/212414a641985684d612/)

Siemens Energy 

### [IMA Sales Leader](https://jobs.opennash.com/companies/api-group/jobs/1e989b134be9f853d234/)

APi Group 

### [Creative and IMA Partner Delivery Manager, gTech Ads Agency and Partners](https://jobs.opennash.com/companies/alphabet/jobs/1db838d979e122385ea5/)

Alphabet · Singapore

OpenNash lists jobs from employers’ own careers sites and does not handle applications.
Dates come from the employer where stated; “first seen” and “last checked” are our own observations.

[ OpenNash ](https://opennash.com/)

AI-native services firm that builds custom software and 24/7 AI agents.

OpenNash Jobs is an OpenNash Applied AI project.
[Book time with OpenNash](https://calendly.com/opennash/new-meeting?hide_gdpr_banner=1)

Jobs
- [Search jobs](https://jobs.opennash.com/search/)
- [Companies](https://jobs.opennash.com/explore/)
- [Hiring report](https://jobs.opennash.com/reports/hiring/)
- [AI assistants](https://jobs.opennash.com/connect/)
- [How it works](https://jobs.opennash.com/methodology/)

OpenNash
- [opennash.com](https://opennash.com/)
- [Services](https://opennash.com/ai-agent-consulting/)
- [OpenNash CX](https://opennash.com/cx/)
- [AI Tools](https://opennash.com/ai-tools/)
- [Blog](https://opennash.com/blog/)
- [About](https://opennash.com/about/)

Contact
- [hello@opennash.com](mailto:hello@opennash.com)

- [LinkedIn](https://www.linkedin.com/company/opennash)

- [Support](https://jobs.opennash.com/support/)

- [Privacy](https://jobs.opennash.com/privacy/)

- [Terms](https://jobs.opennash.com/terms/)

© 2026 OpenNash, Inc. Company names and public job postings belong to their respective owners. Employers control source availability and application decisions.
