# 大模型开发工程师(J11071)

External opening: you cannot apply through AI Native Match; apply on the official site.

Apply through the official channel. The full external receipt stays in your workspace. When signed in, three minimal event types explicitly recorded through the CLI are reported by default: selection, self-reported application and channel observation. Browsing or saving a role does not report an event; external applications do not appear under My applications.

- Company: 云知声 (Listed company)
- Locations: 北京
- Role: Engineering
- Apply on the official site: https://unisound.zhiye.com/social/detail?jobAdId=3da6c16d-a0a1-47c0-b3a0-ea179ee1536d

## Role description from source

1、负责大模型（文本、图像、语音等多模态模型）推理引擎及服务的开发，支持自研及开源模型的快速部署和高性能推理；
2、结合大模型业务场景，优化模型落地过程中的性能问题，降低推理成本。包括但不限于：模型量化蒸馏，服务调度优化等；
3、负责开发和维护大模型MaaS服务，支持高并发调用场景；
4、负责大模型的国产化适配
职位要求
1、硕士及以上学历，1年以上算法相关工程化经验。计算机、电子、自动化、软件等相关专业； 
2、熟悉一种或多种LLM推理框架，有vllm、sglang、lmdeploy、TensorRT-LLM等相关经验优先；
3、熟悉大模型调度优化技术，包括PD分离、多级缓存、多实例调度、负载均衡等，如SGLang model gateway、lm-d；
4、熟悉大模型引擎优化技术，包括prefill cache、tensor并行优化等。有cuda或triton算子优化经验优先；
5、有昇腾、海光、平头哥、沐曦等算力芯片适配经验优先。

---
opening_id: 57c8048d-226f-4d40-a493-3b114429f4b4

`anm openings save --id 57c8048d-226f-4d40-a493-3b114429f4b4`