Ottr is a serverless Public Key Infrastructure framework that handles end-to-end certificate rotations without the use of an agent. The purpose of the blog is to provide an overview on Ottr with sample reference architecture, logical and network flows, and highlight the benefits of the solution. For installation instructions, skip to the Open Source section of the article.

Part two of a series on how we provide powerful, automated, and scalable data privacy and security engineering capabilities at Airbnb.

关于我们在 Airbnb 爱彼迎打造 Android 自动化测试框架的系列文章，一共分为 7 个部分 —— 在第 1 部分中，我们将介绍自动化测试的原理，以及我们做所有自动化测试的基础：状态模拟系统。

Learn more about how to migrate your Kafka cluster from one Zookeeper cluster to another without any user impact.

How innovation in technology and people processes have enabled iOS developers to remain productive in a large codebase.

Consolidating long-running, lightweight tasks for improved resource utilization.

在过去的几年里，Airbnb 爱彼迎的工程团队从单体 Ruby on Rails 架构转变成为了 SOA 面向服务的架构。在我们的 Rails 架构中，每个资源（resource）都有一个单独的 API 来访问底层数据， 这些 API 都有授权检查来保护敏感数据。当时由于访问资源数据的方式单一，因此管理这些授权检查很容易。在向 SOA 过渡的过程中，我们转向了分层架构，其中有封装数据库的数据服务（data services），也有从多个数据服务提取的展示层服务（presentation services）。将权限检查从单体转移到 SOA 的最初方法是将这些检查转移到展示层服务。

Part-III: Building a coherent consumption experience.

Since no one tool was meeting our needs, we decided to build a data protection platform to enable and empower Airbnb to protect data in compliance with global regulations and security requirements. However, in order to protect the data, we first needed to understand it and its associated security and privacy risks.

净推荐值（Net Promoter Score，以下简称 NPS）是 ToC 企业中应用得最广泛的用户满意度衡量指标。通过分析“客户是否愿意向朋友推荐我们的产品”，我们可以检测到用户对企业的认同度与未来的留存率。在爱彼迎，我们同样长期关注这一指标，用以衡量我们为房东、房客社区提供的客户服务品质。

然而在产品快速迭代的今天，传统的 NPS 指标也显现出自身的一些不足。由于 NPS 分数依赖于用户对问卷的反馈，它会不可避免地引入选择性偏差、数据延迟、低覆盖率等问题。这也导致我们几乎无法在短期的 A/B Testing 中使用 NPS 来衡量用户服务端产品的迭代效果。

为了解决这一问题，爱彼迎开发了基于 AI 模型的情绪识别方法，通过分析用户与客服团队对话文本中的情绪变化，来衡量用户此时此刻的满意度状态。在今天的文章中，我们将详细地介绍爱彼迎在情绪识别模型上的开发，以及我们如何一步步将模型生成的结果转化为一个能与企业长期留存率挂钩的高覆盖率、无延迟的新型满意度衡量指标。

Customer Support (CS) can make or break a guest’s travel experience. To support Airbnb’s community of guests and Hosts, we have been investing heavily in developing intelligent CS solutions leveraging state-of-the-art natural language processing (NLP), machine learning (ML), and artificial intelligence (AI) technologies.

In this blog post, we’ll introduce the automated support system at Airbnb, which employs the latest task-oriented conversational AI technology, through the lens of a recently launched feature called Mutual Cancellation. We will describe in detail how we framed the business problem as an AI problem, how we collected and labeled training data, how we designed and built the ML models, and how the models were deployed in the online system. Throughout each step, we’ll discuss some technical challenges we faced during this project and the solutions we innovated to address these challenges.

Gaining trust in data with extensive data quality, accuracy and anomaly checks.

How AI-based Sentiment Models Complement Net Promoter Score.

How Airbnb ships features faster across web, iOS, and Android using a server-driven UI system named Ghost Platform 👻.

Part II: The six design principles of Minerva compute infrastructure.

OMNI是我们用来创建，管理，并通过多种渠道向我们的社区分发内容和信息的平台。OMNI这个名字是指该平台提供全渠道内容的能力。