July 2025 Summaries

Nutrisense, a company that uses personal data to provide actionable health insights, improved its observability and monitoring practices by switching from Datadog to Better Stack, significantly reducing logging expenses. Their backend, built with Ruby on Rails and reliant on GraphQL, faced challenges with logging detail and cost, leading to a necessity for better solutions. The Nutrisense team developed a custom logger to capture essential contextual information, such as GraphQL operation names and user data, and integrated Better Stack to manage these enriched, structured logs efficiently. This transition allowed Nutrisense to perform more effective debugging and tracing, create custom dashboards for real-time performance tracking, and set up targeted alerts for anomalies. Consequently, Nutrisense achieved a scalable, cost-effective observability setup that met their needs without compromise.

The latest changelog from Better Stack Team, updated on July 23, 2025, highlights significant enhancements to dashboard alerts, featuring improved anomaly detection and the ability to customize alerts across multiple applications and platforms. Users can now apply a single alert to various sources without duplication and benefit from more accurate anomaly detection that adapts to both short-term and long-term data patterns, reducing false positives. Additionally, the update includes community content discussing topics such as a comparison between Rust and Go programming languages, monitoring Docker containers, the emergence of the Kimi K2 AI model, and Amazon's new AI IDE, Kiro, which emphasizes spec-driven development. The content is shared under a Creative Commons license, inviting user feedback and article suggestions.

WindBorne Systems is revolutionizing weather forecasting by deploying a global fleet of autonomous, long-duration balloons in the stratosphere to gather extensive atmospheric data from remote regions, thereby enhancing the accuracy of weather models. The company's technology, which has outperformed similar initiatives by Google, is utilized by major organizations such as the European Centre for Medium-Range Weather Forecasts and the US Air Force, contributing to more accurate weather predictions, as demonstrated by a 15% improvement in forecasts for Hurricane Ian. To manage their complex, 24/7 operation, WindBorne implemented Better Stack integrated with Render, providing a unified platform for monitoring logs and metrics. This integration enables real-time debugging and incident resolution, essential for maintaining the fleet's operational reliability and ensuring continuous data collection. The unified system allows for immediate investigation and intervention when anomalies occur, ensuring the success of the balloons' data-gathering missions.

Changelog #9 from Better Stack Team, updated on July 22, 2025, announces that enterprise-level data options are now available to all users on the Pay as you go plan, allowing them to choose where their telemetry data is stored, either in specific regions like US East, US West, Germany, and Singapore, or in a custom S3 bucket for enhanced data control and compliance. Additionally, the update highlights fresh community content that includes a guide to six open-source AI coding assistant alternatives to Cursor, a comparison of Pipenv vs Poetry for dependency management, an evaluation of Google's new AI coding tool Gemini CLI against Claude Code for React tasks, and a demonstration of running Docker on Cloudflare to utilize containers for edge tasks like building a silence remover. The work is shared under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.

Listen Notes, a podcast search engine akin to Google for podcasts, significantly reduced its monitoring costs by 90% by replacing the enterprise-grade tools Datadog, PagerDuty, and Atlassian Statuspage with Better Stack. The switch allowed Listen Notes to consolidate infrastructure monitoring, incident management, and status page functionalities into a single, more cost-effective solution, reducing annual expenses from $3,500 to $348. The transition, completed over two weeks, involved shifting from Datadog's metric-based workflow to Better Stack's log-centric model, configuring heartbeat and uptime alerts for incident management, and creating a branded status page with Better Stack's built-in features. The move not only streamlined Listen Notes' operations but also enhanced the developer experience with quicker setup and more accessible customer support, positioning the company to continue meeting its needs efficiently as it grows.

Changelog #8 introduces enhanced features for Better Stack's Live Tail, including pattern-based filtering and service discovery, allowing users to efficiently locate logs by filtering similar log messages and selecting individual services instead of entire sources. This upgrade facilitates error tracking by enabling users to filter logs based on patterns, exclude irrelevant logs, and select specific services for more focused log analysis. Users can customize service discovery settings to automatically categorize logs by service, even without Kubernetes, by configuring unique log fields. The changelog also highlights new community content, featuring guides and alternatives for popular development tools and frameworks, and it is distributed under a Creative Commons license.