# HubSpot The HubSpot Product Blog is where we riff about topics of interest to us from engineering, to UX, to product management and beyond. This is one page of public article previews, not the complete archive. Follow Next page to continue. Summaries are not the original full articles. ## Incident Report: July 22, 2026 DevFeed: [Incident Report: July 22, 2026]() Original publisher: [Read original article]() Author: Karina Devadoss Published: 2026-08-14T17:06:07Z Content type: news Language: en Sources: [HubSpot]() Topics: [incident](), [configuration](), [bug](), [health checks](), [client]() Tags: [2026](), [bug](), [change](), [configuration](), [customers](), [engineering](), [errors](), [europe](), [health-checks](), [hubspot](), [incident](), [memory](), [network](), [performance](), [report](), [services]() ### AI overview HubSpot reports that a configuration change triggered a latent defect in a core CRM service, causing European customers to experience CRM access disruptions for approximately 55 minutes on July 22, 2026. No data was lost, and non-European regions operated without interruption. ### Source excerpt Customers in Europe faced CRM access issues on July 22 due to a configuration change. Learn about the incident, its impact, and our commitment to improved reliability. ## Building Observable Access Control at Scale: A Rule Engine Approach to JITA DevFeed: [Building Observable Access Control at Scale: A Rule Engine Approach to JITA]() Original publisher: [Read original article]() Author: ankeith@hubspot.com (Andrew Keith) Published: 2026-06-30T18:30:42Z Content type: article Language: en Sources: [HubSpot]() Topics: [Access Control](), [hubspot](), [Authorization](), [observability](), [Security]() Tags: [access-control](), [business](), [customer](), [customer-trust](), [data](), [engineering](), [logging](), [metrics](), [observability](), [security]() ### AI overview HubSpot describes the challenges of managing Just-In-Time Access requests at scale. The article explains how a growing collection of nested conditionals made access decisions difficult to audit and maintain, creating a need for observable, rule-based access control. ### Source excerpt The Problem At HubSpot, we have a fundamental tension built into our culture: we want to help customers succeed, which often means giving employees access to customer portals: Engineers to fix bugs, sales engineers assisting with account setup, or customer support troubleshooting issues during live support calls. But we also need to completely lock down customer data -- because a breach or misuse doesn't just violate customer trust, it can expose sensitive business data, create legal liability, and cause reputational damage that's hard to recover from. Access must be granted only for legitimate, verifiable, and auditable reasons. ## Building the AI Retrieval Infrastructure Behind 20 Billion+ Vectors at HubSpot DevFeed: [Building the AI Retrieval Infrastructure Behind 20 Billion+ Vectors at HubSpot]() Original publisher: [Read original article]() Author: Oleg Tereshin & Xin Liu Published: 2026-06-25T18:17:44Z Content type: article Language: en Sources: [HubSpot]() Topics: [Qdrant](), [AI search](), [Artificial Intelligence](), [Embeddings](), [Retrieval-Augmented Generation](), [Open Source](), [systems](), [quantization]() Tags: [ai](), [embeddings](), [engineering](), [latency](), [open-source](), [quantization](), [rag](), [retrieval](), [search](), [vector-database](), [vectors]() ### AI overview HubSpot describes how it built VaaS, a centralized vector storage and search platform using Qdrant, to support semantic search across tens of billions of vectors and many use cases. ### Source excerpt Discover how HubSpot built a scalable AI retrieval infrastructure, managing over 20 billion vectors with Qdrant, to enhance semantic search and support diverse applications. ## HubSpot Incident Report: February 25, 2026 DevFeed: [HubSpot Incident Report: February 25, 2026]() Original publisher: [Read original article]() Author: Shilpa Varma Published: 2026-03-20T17:48:30Z Content type: article Language: en Sources: [HubSpot]() Topics: [incident](), [ui](), [Access Control](), [Authorization](), [configuration]() Tags: [access-control](), [analysis](), [configuration](), [deployment](), [engineering](), [errors](), [health-checks](), [http](), [hubspot](), [incident](), [issue](), [monitoring](), [report](), [user-interface]() ### AI overview HubSpot reports a February 25, 2026 service disruption that blocked access to certain contact, company, order, and project-based workflows in the user interface. Backend workflow automations continued to run normally, and no customer data was lost. The incident resulted from permission scopes being deployed without corresponding production role assignments. ### Source excerpt On February 25, 2026, HubSpot experienced a service disruption affecting user access to specific types of workflows within our user interface. All backend workflow automations continued to execute normally without interruption - no automations were delayed, missed, or lost. We have completed a thorough analysis of this incident and are implementing targeted safeguards to prevent this class of failure from recurring. ## Automated Code Review: The 6-Month Evolution DevFeed: [Automated Code Review: The 6-Month Evolution]() Original publisher: [Read original article]() Author: eadams@hubspot.com (Emily Adams) Published: 2026-03-02T17:46:43Z Content type: article Language: en Sources: [HubSpot]() Topics: [Code review](), [AI-assisted coding](), [Claude Code](), [GitHub](), [Kubernetes](), [Java]() Tags: [agents](), [ai](), [automated](), [claude-code](), [cli](), [code](), [code-review](), [coding](), [coding-agents](), [cycle-time](), [developer-experience](), [engineering](), [engineering-infrastructure](), [github](), [human-review](), [java](), [kubernetes](), [pull-request](), [pull-requests](), [quality](), [workflow]() ### AI overview HubSpot describes a six-month evolution of its automated code review system. The system uses AI to review pull requests, provide feedback grounded in HubSpot-specific norms, and reduce the time engineers wait for feedback. The article says Sidekick reduced feedback time by 90%, with a peak reduction of 99.76% in September, and traces a shift from a Kubernetes-based system powered by Claude Code toward a framework native to HubSpot's Java stack. ### Source excerpt As HubSpot engineers have increasingly started writing more and more of their code with both local and cloud coding agents, we've noticed that now more than ever, code review is taking up a large portion of cycle time on new changes. ## Cloud Coding Agents at HubSpot DevFeed: [Cloud Coding Agents at HubSpot]() Original publisher: [Read original article]() Author: Stephan Lensky Published: 2026-01-23T15:59:04Z Content type: article Language: en Sources: [HubSpot]() Topics: [hubspot](), [AI-assisted coding](), [Claude Code](), [Self-hosted](), [GitHub](), [pull-requests](), [developer tooling](), [Code review]() Tags: [ai](), [claude-code](), [code-review](), [coding-agents](), [developer-tooling](), [engineering](), [github](), [hubspot](), [self-hosting]() ### AI overview This article explains how HubSpot built and deployed a self-hosted platform for cloud coding agents, integrating them with internal developer tooling, infrastructure, and GitHub. It describes their use of autonomous agents for planning, implementation, and code review, including more than 7,000 merged AI-generated pull requests and over 50,000 human-authored pull requests reviewed by agents. ### Source excerpt This post is the second in a series about empowering product, UX, and engineering teams with AI, especially in the context of writing code. Read the first post about scaling AI adoption here! ## Migrating 76,000 Tests from Enzyme to React Testing Library DevFeed: [Migrating 76,000 Tests from Enzyme to React Testing Library]() Original publisher: [Read original article]() Author: Charley Pugmire Published: 2025-11-13T18:31:47Z Content type: tutorial Language: en Sources: [HubSpot]() Topics: [testing-library](), [React](), [Testing](), [migration](), [ui-testing]() Tags: [engineering](), [migration](), [react](), [testing](), [testing-library](), [ui-testing]() ### AI overview The article explains how a team migrated more than 76,000 tests from Enzyme to React Testing Library over two and a half years. It describes the shift toward user-focused UI tests, the planning process, and the limits of using static analysis alone to estimate the work. ### Source excerpt For years, Enzyme was the de facto standard for testing React applications. Originally built by Airbnb in 2015, Enzyme makes it easy to assert, manipulate, and traverse the output of React components. The last official release of Enzyme came after React 16 was launched in 2017, but even then did not fully support all React 16 features. Ongoing support for Enzyme shifted to the open source community with React 17. However, the creator of the community-maintained Enzyme adapter announced that he would not be offering an updated version for React 18. This effectively marked the end of the Enzyme project, as it would prevent users from leveraging new features like React 18's concurrent mode. ## HubSpot Incident Report for October 20, 2025 DevFeed: [HubSpot Incident Report for October 20, 2025]() Original publisher: [Read original article]() Author: Kartik Vishwanath Published: 2025-11-11T17:09:36Z Content type: news Language: en Sources: [HubSpot]() Topics: [incident](), [Amazon Web Services](), [Cloud](), [Resilience](), [Amazon Simple Queue Service (SQS)](), [Amazon EC2](), [AWS IAM](), [DynamoDB]() Tags: [amazon](), [amazon-sqs](), [analysis](), [aws](), [cloud](), [data](), [database](), [distributed](), [dynamodb](), [ec2](), [engineering](), [failover](), [hubspot](), [iam](), [improvements](), [incident](), [message-queue](), [outage](), [recovery](), [report](), [resilience]() ### AI overview HubSpot reports that a severe AWS outage in the us-east-1 region on October 20, 2025 disrupted multiple product features and affected third-party vendors. The incident caused failures involving DynamoDB, IAM, SQS, and EC2, and degraded HubSpot's TQ2 background task processing. HubSpot says it completed an analysis and is implementing resilience improvements. ### Source excerpt On October 20, 2025, HubSpot experienced a significant service disruption affecting multiple product features due to a severe AWS outage in the us-east-1 region. While our infrastructure remained intact, the widespread nature of the cloud provider failure impacted both our services and critical third-party vendors we rely on. We've completed a thorough analysis of this incident and are implementing comprehensive improvements to strengthen our resilience against future cloud provider disruptions. ## A hardware bug caused intermittent HBase slowdowns at HubSpot DevFeed: [A hardware bug caused intermittent HBase slowdowns at HubSpot]() Original publisher: [Read original article]() Author: Charles Connell Published: 2025-09-24T16:38:06Z Content type: article Language: en Sources: [HubSpot]() Topics: [bug](), [Hardware](), [Databases](), [Java](), [Latency](), [benchmarking](), [Monitoring]() Tags: [2026](), [bug](), [database](), [engineering](), [hardware](), [latency](), [monitoring](), [openjdk](), [performance]() ### AI overview HubSpot investigated intermittent slowdowns in HBase clusters storing CRM data. Increased request-handler occupancy, request latency, disk-read latency, CPU usage, and garbage-collection time made some page loads noticeably slower without triggering automatic failure monitoring. An OpenJDK mitigation is expected in OpenJDK 27. ### Source excerpt Edit, April 30 2026: The OpenJDK now contains a mitigation for the hardware bug described in this article, which will appear in OpenJDK 27. Thank you to Evgeny Astigeevich at AWS for his work on the mitigation. Most software developers have heard the maxim, "it's never a compiler bug," and its companion, "it's never a hardware bug." But if you investigate enough bugs, you'll eventually find an exception to this rule, and that's what we found recently at HubSpot. I discovered how a hardware bug was affecting our users' experience. ## Context is Key: How HubSpot Scaled AI Adoption DevFeed: [Context is Key: How HubSpot Scaled AI Adoption]() Original publisher: [Read original article]() Author: blamattina@hubspot.com (Brian LaMattina) Published: 2025-09-24T14:41:07Z Content type: article Language: en Sources: [HubSpot]() Topics: [AI Strategy](), [AI-assisted coding](), [GitHub Copilot](), [hubspot](), [User experience (UX)](), [Architecture & Design]() Tags: [adoption](), [ai](), [ai-adoption](), [ai-coding](), [engineering](), [github-copilot](), [measurement](), [organizational](), [teams](), [training]() ### AI overview HubSpot describes how it scaled AI coding-tool adoption across its engineering organization, moving from cautious experimentation to broad adoption. The article focuses on organizational commitment, enablement, measurement, and lessons for teams scaling beyond proof-of-concept projects. ### Source excerpt This post is intended to be the first in a series about empowering product, UX, and engineering teams with AI. We're going to focus on how we've approached and scaled the use of AI in the context of writing code.