Airbnb's observability engineering team has transitioned from a legacy StatsD and proprietary Veneur-based aggregation pipeline to a modern, open-source stack utilizing OpenTelemetry Protocol (OTLP), the OpenTelemetry Collector, and VictoriaMetrics' vmagent. The new system handles over 100 million samples per second in production while reducing costs by roughly an order of magnitude.
Key technical highlights include:
* Migration strategy using dual-emitting metrics to bridge legacy StatsD libraries with OTLP adoption.
* Performance improvements, including a reduction in JVM CPU time spent on metrics processing from 10% to under 1%.
* Use of vmagent for streaming aggregation and horizontal sharding to manage high-cardinality data.
* Implementation of a zero injection technique within the vmagent tier to solve Prometheus counter reset edge cases.
* A two-layer architecture consisting of stateless router pods and stateful aggregator pods.
Key technical highlights include:
* Migration strategy using dual-emitting metrics to bridge legacy StatsD libraries with OTLP adoption.
* Performance improvements, including a reduction in JVM CPU time spent on metrics processing from 10% to under 1%.
* Use of vmagent for streaming aggregation and horizontal sharding to manage high-cardinality data.
* Implementation of a zero injection technique within the vmagent tier to solve Prometheus counter reset edge cases.
* A two-layer architecture consisting of stateless router pods and stateful aggregator pods.
Prove AI is developing an observability-first foundation designed for production generative AI systems. Their mission is to enable engineering teams to understand, diagnose, and remediate failures within complex AI pipelines, including LLM inference, retrieval processes, and agent orchestration.
The current release, v0.1, provides an opinionated observability pipeline specifically for generative AI workloads through:
- A containerized, OpenTelemetry-based telemetry pipeline.
- Preconfigured collection of traces, metrics, and logs tailored for AI systems.
- Instrumentation patterns for RAG pipelines, embeddings, LLM inference, and agent-based systems.
- Compatibility with standard backends like Prometheus.
The current release, v0.1, provides an opinionated observability pipeline specifically for generative AI workloads through:
- A containerized, OpenTelemetry-based telemetry pipeline.
- Preconfigured collection of traces, metrics, and logs tailored for AI systems.
- Instrumentation patterns for RAG pipelines, embeddings, LLM inference, and agent-based systems.
- Compatibility with standard backends like Prometheus.
A Python package designed to provide production-ready templates for Generative AI agents on Google Cloud. It allows developers to focus on agent logic by automating the surrounding infrastructure, including CI/CD pipelines, observability, security, and deployment via Cloud Run or Agent Engine.
Key features and offerings include:
- Pre-built agent templates such as ReAct, RAG (Retrieval-Augmented Generation), multi-agent systems, and real-time multimodal agents using Gemini.
- Automated CI/CD integration with Google Cloud Build and GitHub Actions.
- Data pipelines for RAG using Terraform, supporting Vertex AI Search and Vector Search.
- Support for various frameworks including Google's Agent Development Kit (ADK) and LangGraph.
- Integration with the Gemini CLI for architectural guidance directly in the terminal.
Key features and offerings include:
- Pre-built agent templates such as ReAct, RAG (Retrieval-Augmented Generation), multi-agent systems, and real-time multimodal agents using Gemini.
- Automated CI/CD integration with Google Cloud Build and GitHub Actions.
- Data pipelines for RAG using Terraform, supporting Vertex AI Search and Vector Search.
- Support for various frameworks including Google's Agent Development Kit (ADK) and LangGraph.
- Integration with the Gemini CLI for architectural guidance directly in the terminal.
Infinite Monitor is an AI-powered dashboard builder that allows users to describe the widget they want in plain English, and an AI agent will write, build, and deploy it in real time. Each widget is a full React app running in an isolated iframe, offering flexibility and customization. Users can drag, resize, and organize these widgets on an infinite canvas for various applications like cybersecurity, OSINT, trading, and prediction markets.
The project supports multiple AI providers and offers features like dashboard awareness, live web search, and a widget marketplace. It prioritizes security with local-first storage and threat scanning.
The project supports multiple AI providers and offers features like dashboard awareness, live web search, and a widget marketplace. It prioritizes security with local-first storage and threat scanning.
"Prove AI is a self-hosted solution designed to accelerate GenAI performance monitoring. It allows AI engineers to capture, customize, and monitor GenAI metrics on their own terms, without vendor lock-in. Built on OpenTelemetry, Prove AI connects to existing OpenTelemetry pipelines and surfaces meaningful metrics quickly.
Key features include a unified web-based interface for consolidating performance metrics like token throughput, latency distributions, and service health. It enables faster debugging, improved time-to-metric, and better measurement of GenAI ROI. The platform is open-source, free to deploy, and offers full control over telemetry data."
Key features include a unified web-based interface for consolidating performance metrics like token throughput, latency distributions, and service health. It enables faster debugging, improved time-to-metric, and better measurement of GenAI ROI. The platform is open-source, free to deploy, and offers full control over telemetry data."
Distributed tracing is crucial for modern observability, offering richer context than logs. However, the volume of tracing data can be overwhelming. Sampling addresses this by selectively retaining data, with two main approaches: head sampling (deciding upfront) and tail sampling (deciding after collecting all spans). Head sampling is simpler but can miss localized issues. Tail sampling, while more accurate, is complex to implement at scale, requiring buffering, stateful processing, and potentially impacting system resilience. Furthermore, sampling inherently affects the accuracy of RED metrics (request rate, error rate, duration), necessitating metric materialization *before* sampling.
This article details building end-to-end observability for LLM applications using FastAPI and OpenTelemetry. It emphasizes a code-first approach, manually designing traces, spans, and semantic attributes to capture the full lifecycle of LLM-powered requests. The guide advocates for a structured approach to tracing RAG workflows, focusing on clear span boundaries, safe metadata capture (hashing prompts/responses), token usage tracking, and integration with observability backends like Jaeger, Grafana Tempo, or specialized LLM platforms. It highlights the importance of understanding LLM behavior beyond traditional infrastructure metrics.
This article explores the emerging category of AI-powered operations agents, comparing AI DevOps engineers and AI SRE agents, how cloud providers are responding, and what engineers should consider when evaluating these tools.
Logs, metrics, and traces aren't enough. AI apps require visibility into prompts and completions to track everything from security risks to hallucinations.
This post introduces **GIST (Greedy Independent Set Thresholding)**, a new algorithm for selecting diverse and useful data subsets for machine learning. GIST tackles the NP-hard problem of balancing diversity (minimizing redundancy) and utility (relevance to the task) in large datasets.
**Key points:**
* **Approach:** GIST prioritizes minimum distance between selected data points (diversity) then uses a greedy algorithm to approximate the highest-utility subset within that constraint, testing various distance thresholds.
* **Guarantee:** GIST is guaranteed to find a subset with at least half the value of the optimal solution.
* **Performance:** Experiments demonstrate GIST outperforms existing methods (Random, Margin, k-center, Submod) in image classification and single-shot downsampling.
* **Application:** Already used to improve video recommendation diversity at YouTube.
**GIST provides a mathematically grounded and efficient solution for selecting high-quality data subsets for machine learning, crucial as datasets scale.**
.
**Key points:**
* **Approach:** GIST prioritizes minimum distance between selected data points (diversity) then uses a greedy algorithm to approximate the highest-utility subset within that constraint, testing various distance thresholds.
* **Guarantee:** GIST is guaranteed to find a subset with at least half the value of the optimal solution.
* **Performance:** Experiments demonstrate GIST outperforms existing methods (Random, Margin, k-center, Submod) in image classification and single-shot downsampling.
* **Application:** Already used to improve video recommendation diversity at YouTube.
**GIST provides a mathematically grounded and efficient solution for selecting high-quality data subsets for machine learning, crucial as datasets scale.**
.
