Observability for Next-Gen AI Applications: How Automation Anywhere Scaled with Kloudfuse
Executive Summary
Automation Anywhere, a global leader in agentic process automation, faced multiple challenges with its existing observability stack. High costs, fragmented tools, and limited data retention were impeding visibility and slowing down teams. As observability became central to delivering reliable, AI-powered automation at scale, the company needed a unified, scalable, and cost-efficient solution.
By switching to Kloudfuse, Automation Anywhere achieved:
Cut Mean Time to Resolution (MTTR) by
86%
Reduced Observability Costs by
50%
Increased Daily Data Ingestion by
5X
About Automation Anywhere
Automation Anywhere is a global leader in agentic process automation, powering mission-critical systems for enterprises across multiple countries. The company's cloud-native Agentic Process Automation System enables organisations to automate complex workflows across front- and back-office operations, seamlessly integrating both SaaS and legacy systems. Driven by a mission to make automation accessible to everyone, Automation Anywhere delivers enterprise-grade, AI-powered automation at scale, empowering businesses to liberate their workforce from mundane tasks and focus on higher-order innovation and creativity.
Challenges
As a leader in AI-driven automation, Automation Anywhere powers mission-critical systems across the globe. But as they scaled, their existing observability stack was constrained by:
High Licensing & Operational Costs: Engineers had to filter/trim logs, metrics, and traces due to cost, reducing visibility for root cause analysis.
Fragmented Tooling: Logs, metrics, and traces lived in silos, requiring manual correlation that slowed down troubleshooting.
Limited Retention: Observability data was capped at 30 days, hindering trend analysis, AI model training, and long-term investigations.
Data Residency & Sovereignty: Hosting telemetry on third-party platforms conflicted with the company’s strict data governance, residency, and privacy frameworks.
Raghu Sethuraman
VP of Engineering, Automation Anywhere
“Before, we were constantly trading off cost and visibility. With Kloudfuse, our developers are free to explore and troubleshoot using all the data they need. They have freedom, speed, and data abundance, all in one place.”
The Kloudfuse Difference
Adopting Kloudfuse transformed observability into a strategic enabler.
Unified Observability
Automation Anywhere consolidated metrics and logs in one platform, with distributed tracing coming next, offering a single-pane-of-glass view across environments.
Cost Efficiency + Unlimited Retention
By running Kloudfuse within their own infrastructure, they eliminated vendor markups and gained control over data retention, now keeping logs and metrics for 2–3 years at minimal cost.
AI-Ready Data Lake
With deeper data access, the team can now:
Train predictive models on historical usage patterns
Forecast customer behavior and product usage
Improve the reliability of their products using AI-driven insights
The Kloudfuse data lake also serves as a base layer for MLOps and agentic workflows, enabling:
Seamless monitoring of AI/ML workloads across AWS, Azure, and GCP environments.
Feeding observability data into intelligent, agentic workflows
By hosting data internally, Automation Anywhere is future-proofing its technology. With no rate limits or retention caps, developers building agentic workflows now have complete freedom to explore, analyze, and operationalize data at scale.
Complete Data Ownership
Kloudfuse’s Self-SaaS, BYOC (Bring Your Own Cloud) deployment ensures all observability data stays within Automation Anywhere’s network, meeting strict data privacy, governance, and residency requirements.
Broad Adoption, Developer Freedom
Today, hundreds of users across multiple functions, including Platform, QA, DevOps, Customer Support, and Product, access observability data directly through Kloudfuse's self-service dashboards. The platform powers observability across 15 global regions, monitoring 80–90 Kubernetes clusters deployed across AWS, GCP, and Azure, serving customers globally.
Quantifying Success: Key Advantages Realized with Kloudfuse
Upon migrating to Kloudfuse, Automation Anywhere experienced a significant transformation in its operational efficiency, resulting in substantial returns, including:
86% Reduction in MTTR: By leveraging the Kloudfuse platform, Automation Anywhere has dramatically accelerated incident resolution, reducing MTTR from approximately 3 weeks to 3 days or less. This improvement is driven by Kloudfuse's unified observability platform, which eliminates the need to correlate data across multiple siloed tools and manually match timestamps between logs and metrics.
5X Increase in Data Ingestion: With Kloudfuse, Automation Anywhere scaled from 30TB/day to 150TB/day of telemetry data, a 5X increase. Engineers no longer need to filter out debug logs or trim metrics to manage costs, enabling complete visibility for root cause analysis.
~50% Reduction in Total Observability Costs: Despite handling 5X more data, Automation Anywhere achieved approximately 50% overall savings, with significantly lower license and storage costs.
Unlimited Data Retention: Data retention expanded from 30 days to unlimited, with the ability to hydrate archived logs from S3 for compliance, legal, or historical investigations.
Broad User Access: Platform access expanded from a centralized team to hundreds of users across Engineering, DevOps, QA, Customer Support, and Product teams—giving engineers the freedom to query without restrictions.
Full Data Ownership & Residency: With Kloudfuse's Self-SaaS deployment model, Automation Anywhere now hosts all observability data internally, gaining complete data ownership while meeting corporate data residency and governance policies.
Why Automation Anywhere Chose Kloudfuse for Observability Transformation
Flat Pricing Model: Kloudfuse’s transparent licensing allows developers to troubleshoot freely, without worrying about ingest, query, or retention-based overages.
End-to-End Visibility: Unified logs, metrics, and tracing within a single integrated data lake provides full observability across the stack, accelerating root cause analysis and troubleshooting.
Freedom of Exploration for Developers & AI Teams: No need to spend time trimming or filtering data. Engineers can explore unfiltered, full-fidelity datasets for faster root cause analysis and better product development.
Built-in fingerprinting and auto facet extraction automatically identify and index key log attributes, eliminating manual parsing.
Full support for high-cardinality data without limits.
Uses an industry-standard query language, not vendor-specific syntax, minimizing the learning curve.
Multi-Agent Ingestion Support: Seamless integration with a variety of agents, including Datadog Agent and OpenTelemetry (OTel), ensures flexibility across environments.
Cloud-Agnostic: Centralized observability across AWS, GCP, and Azure, with deep support for large-scale Kubernetes environments across 15 regions.
Data Ownership & BYOC Deployment: Kloudfuse runs within Automation Anywhere’s own cloud environment, keeping sensitive AI data private and compliant with internal data governance, residency, and security policies.
Fast, Personalized Support: Feature requests are typically turned around in days, not months, offering what the team calls the best support experience they’ve had.
“The ability to ingest massive amounts of data and set up alerts without worrying about cost or rate limits is a game changer. Combined with Kloudfuse’s ability to handle high-cardinality data and a wide variety of data types, from traditional MELT data to MLOps telemetry from AI/ML workloads, it ensures we don’t miss anything critical.”
Sr Staff Engineer Engineer, Automation Anywhere
Looking Ahead
Automation Anywhere is expanding its use of Application Performance Monitoring (APM) and Large Language Model (LLM) observability within the Kloudfuse platform to consolidate tooling, reduce complexity, and provide developers with a unified source of truth across all telemetry data.
They are also focusing on monitoring the next generation of applications, including AI/ML workloads, where data from various AI and MLOps platforms is fed into Kloudfuse for advanced observability.
Conclusion
By moving to Kloudfuse, Automation Anywhere now benefits from unified, real-time visibility across its observability data, alongside improved cost efficiency, agility, and control. The platform directly supports the company’s mission to deliver enterprise-grade, AI-powered automation at a global scale.
With 5X more data, 86% faster resolution times, and ~50% lower observability costs, Kloudfuse has become a foundational pillar for Automation Anywhere’s reliability, speed, and innovation. Today, hundreds of users across engineering, support, and product teams rely on Kloudfuse to power data-driven decisions and ensure seamless performance across global environments.
Kloudfuse has unified our metrics, logs, and traces into a single observability platform, providing access to more data than we had before. This transformation empowers our developers with the freedom and speed to troubleshoot issues at scale today, while providing a base layer of abundant data for monitoring the new generation of applications we are building, enabling us to confidently scale our AI products globally.
VP of Engineering, Automation Anywhere