社区帖子招聘【项目制/全职远程】Senior Python / LLM Pipeline Engineer — AI Elder Care Platform | 高级LLM管道工程师

关于项目 / About the project
Oakcare 是一家新加坡养老护理平台，帮助家庭雇佣可信赖的住家护理师。我们的系统基于 Google ADK + Gemini 2.5（GCP），目前有可运行的代码库，需要一位高级工程师在扩大用户规模前稳定 AI 管道。

Oakcare is a Singapore-based platform that helps families hire trusted live-in eldercare workers. We use AI agents (Google ADK + Gemini 2.5 on GCP). We have a working codebase and need a senior engineer to stabilise the AI pipeline before we scale to customers.

你将负责 / What you'll work on

• 修复静默失败问题——实现明确的成功/失败状态 / Fix silent failure patterns — implement explicit success/failure states
• 解决导致约40%工作经历数据丢失的竞态条件 / Resolve race conditions causing ~40% data loss in work experience pipeline
• 为所有 LLM 输出实现 JSON schema 校验 / Implement JSON schema validation for all LLM outputs
• 为所有 LLM API 调用添加指数退避重试逻辑 / Add retry logic with exponential backoff for LLM API calls
• 修复 WhatsApp 聊天机器人约束处理和降级搜索问题 / Fix WhatsApp chatbot constraint handling and fallback search
• 修复 Resume Boost 管道日期解析、评分确定性和过时数据问题 / Fix Resume Boost pipeline — date parsing, scoring, stale data

技术要求 / Requirements

• 扎实的 Python，熟悉异步模式 / Strong Python — async patterns, state management
• LLM 智能体框架实战（Google ADK、LangChain 或同类）/ LLM agent framework experience — Google ADK, LangChain, or similar
• 生产环境 LLM API 集成经验 / Production LLM API integration (Gemini, OpenAI, or similar)
• GCP、Cloud Run 经验
• 加分：Twilio、Xano、Firecrawl

薪资 / Compensation：面议 / Negotiable — 项目制 / Project-based

申请时请附上相关项目经验或代码样本。
Please include relevant project experience or code samples showing LLM pipeline work.