人工智能与全栈开发工程师
AI & Full-Stack Developer
位置：南非境内远程办公
班次：美国工作时间
雇佣类型：每周20小时
独立承包商
职位概述
我们正在寻找一位经验丰富的AI聚焦型全栈开发人员，负责构建和部署智能软件应用。在此职位中，您将结合传统的全栈Web工程与现代AI能力——构建定制的RAG管道、部署AI代理工作流、连接LLM API，并将AI驱动的自动化嵌入到运营平台中。
您将直接与我们的团队合作，将业务流程转化为快速、可扩展的AI驱动网络应用。
主要职责
- AI与LLM集成：使用OpenAI（GPT-4o）、Anthropic（Claude）、开源LLM和API集成设计并实现智能功能。
- RAG与代理工作流：使用LangChain、LlamaIndex或CrewAI等框架构建检索增强生成（RAG）系统、自主AI代理和自定义工作流自动化。
- 数据库与向量搜索：设计、管理并查询关系型数据库（PostgreSQL、MySQL）和向量数据库（Pinecone、Qdrant、Supabase pgvector），用于语义文档检索和记忆存储。
- 全栈应用工程：开发响应式前端界面（React、Next.js、Vue）和稳健的后端API（Python FastAPI/Flask、Node.js或TypeScript）。
- 优化与性能：实施有效的提示工程策略，评估模型延迟和API成本，并优化数据库查询以确保高可靠性。
- 每日敏捷执行：参与每日站会，维护清晰的Git工作流程，编写自动化单元测试，并通过Slack和Jira/Asana提供定期进度更新。
所需技术技能与经验
- 专业经验：4年以上软件工程经验，至少1.5年以上使用AI模型和LLM构建生产应用的实际经验。
- AI与机器学习生态系统：
o 框架：有使用LangChain、LlamaIndex或原始LLM API工具调用/函数执行的经验。
o 向量存储：有使用Pinecone、Chroma、Supabase（pgvector）或Weaviate的经验。
o 提示与微调：深入理解提示工程、函数调用、代理循环和模型评估。
- 核心技术栈：
o 后端：精通Python（FastAPI、Django）或TypeScript/Node.js。
o 前端：React.js、Next.js或Vue.js。
o 版本控制与云：Git、GitHub/GitLab、AWS、Azure或Google Cloud。
查看英文原文
Location: Remote, within South Africa
Shift: U.S. Business Hours
Employment Type: 20 Hours/Week
Independent Contractor
Position Overview
We are seeking an experienced AI-Focused Full-Stack Developer based to architect, build, and deploy intelligent software applications. In this role, you will combine traditional full-stack web engineering with modern AI capabilities—building custom RAG pipelines, deploying AI agentic workflows, connecting LLM APIs, and embedding AI-driven automation into operational platforms.
You will collaborate directly with our team to turn business workflows into fast, scalable, AI-powered web applications.
Key Responsibilities
- AI & LLM Integration: Design and implement intelligent features using OpenAI (GPT-4o), Anthropic (Claude), open-source LLMs, and API integrations.
- RAG & Agentic Workflows: Build Retrieval-Augmented Generation (RAG) systems, autonomous AI agents, and custom workflow automations using frameworks like LangChain, LlamaIndex, or CrewAI.
- Database & Vector Search: Design, manage, and query relational databases (PostgreSQL, MySQL) and vector databases (Pinecone, Qdrant, Supabase pgvector) for semantic document retrieval and memory stores.
- Full-Stack Application Engineering: Develop responsive front-end interfaces (React, Next.js, or Vue) and robust back-end APIs (Python FastAPI/Flask, Node.js, or TypeScript).
- Optimization & Performance: Implement effective prompt engineering strategies, evaluate model latency and API costs, and optimize database queries for high reliability.
- Daily Agile Execution: Participate in daily stand-ups, maintain clean Git workflows, write automated unit tests, and provide regular progress updates via Slack and Jira/Asana.
Required Technical Skills & Experience
- Professional Experience: 4+ years in software engineering, with at least 1.5+ years of hands-on experience building production applications using AI models and LLMs.
- AI & Machine Learning Ecosystem:
o Frameworks: Hands-on experience with LangChain, LlamaIndex, or raw LLM API tool calling/function execution.
o Vector Stores: Experience working with Pinecone, Chroma, Supabase (pgvector), or Weaviate.
o Prompting & Fine-Tuning: Deep understanding of prompt engineering, function calling, agentic loops, and model evaluation.
- Core Tech Stack:
o Back-End: Strong proficiency in Python (FastAPI, Django) or TypeScript/Node.js.
o Front-End: React.js, Next.js, or Vue.js.
o Version Control & Cloud: Git, GitHub/GitLab, Vercel, Docker, AWS, or GCP.
Originally posted on Himalayas