Job Description

嵌入式操作系统开发（AI Agent方向）-资深工程师/专家

上海

社招

全职

数字技术 - 数字安全

本科及以上

10 年以上

职位描述

1. 从车载嵌入式操作系统角度建立 Agent 资源使用监控、资源画像和趋势评估机制，覆盖 CPU、内存、IPC、IO、GPU/NPU 队列、容器/cgroup 等关键资源。 Establish OS-level mechanisms in an in-vehicle embedded environment for monitoring, profiling, and trend assessment of Agent resource usage, covering CPU, memory, IPC, IO, GPU/NPU queues, containers/cgroups, and other critical resources. 2. 设计车载资源受限和实时约束场景下，单一共享 LLM、多业务 Agent 共存的任务准入、资源隔离、deadline 保护、推理降级和执行中重评估机制。 Design task admission, resource isolation, deadline protection, inference degradation, and runtime re-evaluation mechanisms for multiple business agents sharing a single LLM under in-vehicle resource-constrained and real-time constraints. 3. 负责 Agent 工具调用链到车端服务的适配、权限控制、超时处理、错误码映射、撤权和审计机制设计与开发，支撑跨域控制器通信和车载 SOA 服务访问。 Design and develop adaptation, permission control, timeout handling, error-code mapping, revocation, and audit mechanisms from Agent tool invocation chains to in-vehicle services, supporting cross-domain controller communication and in-vehicle SOA service access. 4. 负责车载多 Agent 场景下的运行状态观测、异常检测和系统级故障定位，支撑任务风暴、协同停滞、Agent 死锁、工具链失控和共享模型资源竞争等问题分析。 Build runtime observability, anomaly detection, and system-level fault diagnosis capabilities for in-vehicle multi-agent scenarios, supporting analysis of task storms, collaboration stalls, agent deadlocks, tool-chain runaway, and shared model resource contention.

职位要求

1. 本科及以上学历，计算机、软件工程、自动化、电子工程、控制工程或相关专业。 Bachelor’s degree or above in Computer Science, Software Engineering, Automation, Electronic Engineering, Control Engineering or related fields. 2. 具备车载嵌入式 OS 系统软件开发经验，参与过资源管理、调度优化、IPC 通信、容器/cgroup 隔离、系统监控、故障诊断或性能优化中的至少两类工作。 Experience in in-vehicle embedded OS system software development, with hands-on involvement in at least two areas such as resource management, scheduling optimization, IPC communication, container/cgroup isolation, system monitoring, fault diagnosis, or performance optimization. 3. 具备运行时资源观测与系统级问题定位能力，能够基于日志、trace、性能计数器、调度延迟、队列深度、内存/IO/IPC 指标等信息分析资源竞争、阻塞、超时和性能退化问题。 Strong capability in runtime resource observability and system-level diagnosis, with the ability to analyze resource contention, blocking, timeout, and performance degradation issues based on logs, traces, performance counters, scheduling latency, queue depth, memory, IO, and IPC metrics. 4. 具备扎实的 C/C++ 或 Rust 开发能力，熟悉嵌入式或系统级软件开发方法，能够完成操作系统相关模块、运行时监控组件、通信适配组件或资源管理机制的设计与实现。 Strong development capability in C/C++ or Rust, familiar with embedded or system-level software development practices, and able to design and implement OS-related modules, runtime monitoring components, communication adaptation components, or resource management mechanisms. 5. 了解 LLM 部署、Agent Runtime、Tool Calling、模型服务或云端 AI Infra 中至少一种，有实际接触经验。 Familiar with at least one of LLM deployment, Agent Runtime, Tool Calling, model serving, or cloud AI infrastructure, with hands-on exposure preferred.

投递