AI工程师 - 仅限股权激励
AI Engineer - Equity Only
职位名称：AI 工程师
关于 Luupli
Luupli 是一款以股权、多样性和平等为核心的社会媒体应用。我们相信社交媒体可以成为一种善的力量，我们致力于打造一个能最大化创作者和企业从中获得价值的平台，同时对社会和地球产生积极影响。我们的应用目前处于测试阶段，我们对它带来的可能性感到兴奋。我们的团队由充满热情和奉献精神的个人组成，他们致力于让 Luupli 取得成功。
职位描述
作为 Luupli 的 AI 工程师，你将在开发智能系统和编排代理工作流方面发挥关键作用，这些工作流驱动 Luupli 的 AI 功能。你的工作将涵盖检索增强生成（RAG）、多代理 LLM 编排、自动字幕、生成式媒体和内容审核。
你将使用 LangGraph、LangChain 和 Google 的 Agent Development Kit 等框架，在 Google 云平台上构建持久、可扩展的 AI 服务。这是一个全栈 AI 职位，涵盖智能后端 API、LLM 代理编排以及与面向产品功能的集成。
职责
- 使用 LangGraph、LangChain 或 Google 的 Agent Development Kit 构建和部署多代理 AI 工作流。
- 使用嵌入、语义分块和向量数据库（如 FAISS、Pinecone、Weaviate）实现 RAG 流水线。
- 将托管和开源 LLM（OpenAI、Gemini、Claude、Ollama、Mistral）集成到智能系统中。
- 使用 FastAPI 构建 REST API，并使用 Streamlit 构建内部工具以暴露 AI 功能。
- 在 GCP 上使用 Vertex AI、Cloud Run、Cloud Functions、IAM 和 Pub/Sub 部署生产级服务。
- 将 AI 嵌入到平台功能中，例如自动字幕、LuupForge（生成式工作室）、信息流个性化和实时审核。
- 在 AI 系统生命周期中维护模块化、可测试、可观测和安全的代码。
要求
- 3 年以上应用 AI/ML 工程经验（生产级部署，而非仅限于研究）。
- 强大的 Python 开发技能，具备全栈 AI 工程经验：
- FastAPI、Streamlit
- LangGraph、LangChain 或类似框架
- PyTorch、Transformers
- FAISS、Weaviate 或 Pinecone
- 具有使用托管 API（OpenAI、Gemini）和自托管模型（Mistral、Ollama、LLaMA）的丰富经验。
- 对 LLM 编排、代理工具使用、记忆共享和提示工程有深入理解。
- 具备实际操作经验
查看英文原文
Job Title: AI Engineer
About Luupli
Luupli is a social media app that has equity, diversity, and equality at its heart. We believe that social media can be a force for good, and we are committed to creating a platform that maximizes the value that creators and businesses can gain from it, while making a positive impact on society and the planet. Our app is currently in Beta Test, and we are excited about the possibilities it presents. Our team is made up of passionate and dedicated individuals who are committed to making Luupli a success.
Job Description
As an AI Engineer at Luupli, you will play a pivotal role in developing intelligent systems and orchestrating agentic workflows that power Luupli’s AI features. Your work will span Retrieval-Augmented Generation (RAG), multi-agent LLM orchestration, auto-captioning, generative media, and content moderation.
You’ll use frameworks like LangGraph, LangChain, and Google’s Agent Development Kit to build persistent, scalable AI services on Google Cloud Platform (GCP). This is a full-stack AI role that spans intelligent backend APIs, LLM agent orchestration, and integration with product-facing features.
Responsibilities
- Build and deploy multi-agent AI workflows using LangGraph, LangChain, or Google’s Agent Development Kit.
- Implement RAG pipelines using embeddings, semantic chunking, and vector databases (e.g., FAISS, Pinecone, Weaviate).
- Integrate hosted and open-source LLMs (OpenAI, Gemini, Claude, Ollama, Mistral) into intelligent systems.
- Build REST APIs with FastAPI and internal tools with Streamlit to expose AI functionality.
- Deploy production-grade services on GCP using Vertex AI, Cloud Run, Cloud Functions, IAM, and Pub/Sub.
- Embed AI into platform features such as auto-captioning, LuupForge (generative studio), feed personalization, and real-time moderation.
- Maintain modular, testable, observable, and secure code across the AI system lifecycle.
Requirements
- 3+ years experience in applied AI/ML engineering (production-level deployments, not research-only).
- Strong Python development skills with full-stack AI engineering experience:
- FastAPI, Streamlit
- LangGraph, LangChain, or similar
- PyTorch, Transformers
- FAISS, Weaviate, or Pinecone
- Solid experience working with hosted APIs (OpenAI, Gemini) and self-hosted models (Mistral, Ollama, LLaMA).
- Deep understanding of LLM orchestration, agent tool-use, memory sharing, and prompt engineering.
- Hands-on experience with Google Cloud Platform (GCP); especially Vertex AI, Cloud Functions, Cloud Run, and Pub/Sub.
- Familiarity with best practices in cloud-based software development: containerization, CI/CD, testing, monitoring.
Nice to Have
- Experience with Google’s Agent Development Kit or similar agent ecosystems.
- Familiarity with multimodal AI (e.g., handling text, image, audio, or video content).
- Prior experience developing creator platforms, content recommendation engines, or social media analytics.
- Understanding of ethical AI principles, data privacy, and bias mitigation.
- Experience with observability tools (e.g., Sentry, OpenTelemetry, Datadog).
- Data engineering experience, such as:
- Building ETL/ELT pipelines
- Working with event-based ingestion and structured logs (e.g., user sessions, reactions, feeds)
- Using tools like BigQuery, Airflow, or dbt
- Designing or consuming feature stores for AI/ML applications
Compensation
Equity-Only, Purpose-Driven
At Luupli, we believe in building something extraordinary together. This role is equity-only at this stage, meaning you’re not just working for Luupli, you’re building it with us. If you’re driven by impact, excited to shape the future of social media, and ready to share in the upside of what we’re creating, this is your chance to be part of the founding story.
Notice to Recruitment Agencies
Luupli does not accept unsolicited CVs or candidate profiles from recruitment agencies. We do not enter into contracts or pay fees for candidates submitted without a prior written agreement in place. Please note that any CVs received in this manner will be considered the property of Luupli, and no fees will be payable.
Originally posted on Himalayas