July 2025 Summaries

Observe has announced a significant milestone with the closure of $156 million in Series C funding, marking a pivotal moment in its mission to innovate software development for the AI era through advanced observability solutions. Acknowledging the challenges of traditional observability methods, such as siloed tools and overwhelming data volumes, Observe introduces its modern, scalable architecture, including the O11y Data Lake and the O11y Knowledge Graph, to efficiently manage telemetry data in open formats like OpenTelemetry and Apache Iceberg. These innovations aim to reduce costs, enhance data ownership, and improve problem-solving capabilities with the AI-driven O11y AI SRE system. The company has demonstrated impressive growth, tripling revenue and expanding its customer base, with notable contracts replacing legacy vendors like Splunk and New Relic. With the new funding, Observe intends to focus on further research and development of its AI capabilities and scaling its sales and customer success teams, positioning itself as a leader in the observability field by prioritizing customer success and effective data management over traditional models.

Vibe Loop is an emerging model in reliability engineering that emphasizes a proactive, AI-native feedback loop for improving production systems by integrating code writing, observation, learning, and rapid improvement. Unlike traditional systems that are often reactive, Vibe Loop uses AI to automatically generate and improve instrumentation, surface blind spots, and adapt telemetry, allowing developers to work more efficiently by providing insights and actionable suggestions. This concept incorporates "vibe coding" and "vibe ops," where AI aids in tasks traditionally handled by engineers, such as diagnosing issues and suggesting code changes, facilitating a more seamless and continuous workflow without the need for extensive manual intervention or documentation. Although not a standalone tool, Vibe Loop represents a shift towards more intelligent systems that collaborate with engineers to enhance reliability and reduce toil, ultimately offering a dynamic way to handle complex production environments.

Observe has revamped its website to emphasize its capabilities in AI and observability at scale, highlighting its robust infrastructure that can handle vast amounts of telemetry data, ingesting tens or hundreds of terabytes into a single instance and storing it efficiently in an S3-based data lake with Apache Iceberg format for flexibility. The platform boasts exceptional system concurrency, offering over 99.9% availability with options for platinum resiliency, which includes a 15-minute failover to another availability zone. Unlike traditional observability tools, Observe does not rely on building indexes for each time series, allowing for efficient data filtering and sampling without additional tools. The platform supports high query volumes, executing an average of 220 million queries per day and scanning around 16 PiB of data daily, making it suitable for customers with significant scale demands.

The Observe MCP (Model Context Protocol) Server is a hosted gateway that allows AI agents to translate natural language questions into secure OPAL queries on an Observe data lake, providing answers in tables or charts without requiring schema knowledge. By integrating with popular AI tools like Claude Desktop, Cursor, and Augment, the server enhances troubleshooting by converting a complex investigation into a conversational process, enabling workflows that automate root-cause analysis, adaptive log triage, and business-impact summaries. Early feedback from users highlights the system's ability to match UI results, expand query time ranges dynamically, and even propose code changes, significantly reducing incident resolution times and improving efficiency in observability tasks. The server requires minimal setup, operates securely within Observe's SaaS environment, and supports engineers and developers in maintaining focus within their tools, offering a transformative approach to managing and analyzing data.

The post discusses the evolution of product thinking in the context of Observability and software engineering, emphasizing the need to consider the entire product lifecycle, from code development to production operations. It introduces the concept of "Vibe Loop," which leverages "Vibe Coding" and code generation tools to enhance code instrumentation, aiming to standardize and improve the debugging process. The idea is to use AI agents to gather contextual information, thereby expediting root cause analysis and allowing engineers to focus on feature development while maintaining robust instrumentation. The post advocates for a proactive approach to Observability, where AI agents not only identify and address potential issues before they escalate but also recommend and implement improvements, reducing manual efforts and enhancing efficiency in troubleshooting.

As data volumes grow rapidly due to the proliferation of microservices and various teams generating high telemetry, managing log data becomes challenging, often leading to unexpected costs and time-consuming log management efforts. Observe offers a solution with Drop Filters, now generally available, which enables users to discard unwanted log lines before they impact data ingest volume, serving as a temporary relief valve while upstream fixes are implemented. Drop Filters can reduce ingest volumes by up to 50%, allowing teams to better manage costs and avoid labor-intensive log tuning. They are easily configured through Observe's user interface or managed via infrastructure-as-code with Terraform, providing real-time metrics on data filtering effectiveness. The tool aims to streamline log management, reduce surprise overage charges, and facilitate smoother cross-team coordination, ultimately helping organizations maintain budget adherence while dealing with the complexities of modern observability data.

Observe is an AI-powered observability platform that enhances performance, troubleshooting, and cost efficiency by correlating logs, metrics, and traces within an open data lake. It aims to expand access to its capabilities through the Model Context Protocol (MCP) server, which connects AI agents with external resources, allowing them to gather context and generate insights, code changes, or error investigations. Unlike typical MCP implementations that simply wrap APIs, Observe has developed an AI-first approach by incorporating a Knowledge Graph, which indexes infrastructure and business data with enriched metadata to assist AI agents in generating contextually relevant queries. These queries, structured using a JSON schema, are then translated into the OPAL query language to provide complex insights, which are represented visually through markdown-based tables and PNGs. By hosting its MCP server, Observe offers controlled execution, remote configuration, and rapid updates, allowing enterprise customers to integrate both their private data and SaaS data for comprehensive observability.