One platform for complete visibility, control, and AI-ready operations BENGALURU, India, December 03, 2025 - As enterprises race to deploy AI applications, they're discovering that their observability infrastructure can't keep pace. Traditional monitoring tools built for cloud-native workloads don't understand LLM behavior, token consumption, or prompt-response patterns. Adding AI-specific monitoring creates new silos. Observability costs spiral unpredictably, and compliance becomes a maze of fragmented data policies across disconnected platforms. Kloudfuse today launched Kloudfuse 3.5 to solve this fragmentation. The unified observability platform introduces Model Context Protocol (MCP) integration for natural language queries, FIPS 140-2/3 validated cryptographic modules for regulated industries, and comprehensive platform engineering controls, delivering complete visibility across traditional and AI workloads without separate tools or unpredictable costs. "The observability industry created this problem by treating every new technology as a separate product," said Ashish Hanwadikar, CTO and Co-Founder of Kloudfuse. "We took the opposite approach: integrate LLM telemetry directly into APM using OpenTelemetry standards. Platform teams get the same operational control over AI workloads as traditional services. Our MCP server unifies signals across metrics, logs, traces, and events for comprehensive insights through natural language queries, one platform, one instrumentation approach, predictable costs." Since launching Kloudfuse 3.0 in November 2024, Kloudfuse has shipped over 50 major capabilities, proving that unified observability doesn't require fragmented tools, vendor lock-in, or choosing between AI capabilities and cost control. AI-Native Observability: Natural Language Queries Meet LLM Monitoring Kloudfuse 3.5 introduces natural language access to observability data through the Model Context Protocol. Instead of constructing complex queries or navigating multiple dashboards, engineers can ask questions in plain language: "Show me services with elevated error rates in the last hour," or "Which deployments caused memory spikes today?" The MCP server translates these requests into FuseQL, queries across metrics, logs, and traces, and returns comprehensive insights linking application behavior to infrastructure impact. Platform teams can leverage this for automated workflows, building agents that trigger incident responses, analyze capacity trends, or identify cost anomalies. The standards-based implementation works across LLM providers, avoiding vendor lock-in at the AI layer just as OpenTelemetry avoids it at the observability layer. Kloudfuse 3.5 also integrates LLM monitoring directly into APM.. It tracks prompt and completion tokens across providers, monitors model latency and error rates, and correlates AI requests with backend service performance, all through unified OpenTelemetry instrumentation. Teams deploying AI applications get full-stack visibility without adding separate monitoring tools or duplicate agents. Federal Security Certifications for Regulated Industries "For enterprises in regulated sectors, FIPS validation isn't optional, it's contractually required," said Pankaj Thakkar, CEO and Co-Founder of Kloudfuse. "Major observability vendors haven't pursued these certifications, leaving enterprises choosing between comprehensive monitoring and compliance requirements. Kloudfuse 3.5 eliminates that trade-off." Kloudfuse 3.5 implements FIPS 140-2/3 validated cryptographic modules across data ingestion, storage, queries, and API access. Building on FIPS validation, the platform establishes a clear FedRAMP authorization pathway with NIST 800-53 security controls and automated compliance reporting. Combined with VPC deployment, organizations retain complete data sovereignty, as observability data never leaves customer infrastructure. "At Zscaler's scale, observability must be both robust and compliant," said Kishore Thakur, Senior Director, Cloud Platform Engineering at Zscaler. "Kloudfuse handles our massive data volumes while meeting FIPS 140-2 compliance and FedRAMP pathway requirements. For enterprises with strict compliance needs, this combination of flexibility, security, and innovation is exceptional." Data Governance: Managing Access and Compliance Enterprise data governance requires more than encryption. It demands control over who accesses what data, how long it's retained, and how changes are tracked. Kloudfuse 3.5 delivers comprehensive data governance across all telemetry streams with data scrubbing that lets teams preview records before deletion, apply filters targeting specific data, and maintain audit trails for GDPR, HIPAA, and internal compliance requirements. Stream-specific RBAC and identity management implement granular access control through labels and tags. Engineering teams access only their service telemetry. Security teams query across all streams for investigations. Finance teams view cost metrics without exposure to application traces. Automatic synchronization with identity providers like Okta and Google ensures access policies stay current as teams evolve, new hires, role changes, and departures update permissions without manual intervention. Self-ingested audit logging creates queryable compliance trails using FuseQL. Teams can search configuration changes, visualize policy updates over time, and set alerts for unauthorized access attempts, turning compliance documentation into observable data. Platform Engineering: Cost Control and Operational Flexibility Observability at scale demands operational controls, not just monitoring capabilities. Kloudfuse 3.5 introduces granular rate control at the stream level, allowing platform teams to set independent ingestion limits for each telemetry type. Filters within streams ensure critical production data takes priority over verbose debug logs or development environment noise, preventing runaway ingestion costs from misconfigured services. Consumption tracking dashboards break down data volumes and costs by team, service, and environment. Organizations can implement chargeback models showing each team's observability footprint, driving accountability for data generation. Finance teams gain visibility into spending patterns, while platform teams identify optimization opportunities. For reliability, Kloudfuse 3.5 supports multiple availability configurations. Multi-zone deployments provide automatic failover with continuous operations across availability zone failures. Organizations requiring lower resource overhead can deploy disaster recovery configurations with quick manual failover, balancing availability needs against infrastructure costs. Custom Metrics SLOs enable teams to define service level objectives beyond infrastructure metrics. They can track business outcomes like transaction success rates, API quota consumption, or data processing throughput using PromQL, applying SLO discipline to metrics that matter for business operations, not just technical performance. Service accounts with bearer-token authentication support automation workflows, enabling teams to manage observability configuration through infrastructure-as-code practices. Advanced Query Capabilities for Large-Scale Data As observability data grows, query performance becomes a bottleneck. Kloudfuse 3.5 addresses this with scheduled views, precomputed aggregations that refresh at configurable intervals. Complex queries that previously required full table scans now return results instantly from materialized datasets. Automated scheduled searches run periodic analyses: security posture assessments, resource utilization trends, and SLA compliance reports, delivering insights without manual query execution. The platform expands FuseQL's analytical operators for advanced investigations. Teams can compare metrics across time windows to pinpoint deployment impacts, parse nested JSON structures without preprocessing, and enrich telemetry with reference data through lookup tables, mapping service identifiers to cost centers or correlating IP addresses with threat intelligence, at query time rather than ingestion. "The enhanced FuseQL capabilities give our teams powerful new ways to analyze trends faster," said Raghu Sethuraman, Vice President of Engineering at Automation Anywhere. "The MCP server enables our developers to interact with observability data in entirely new ways. Complete ownership through Self-SaaS deployment is what we need for the next generation of AI-powered automation." Open Standards: Avoiding Vendor Lock-In Kloudfuse's architecture relies on open standards rather than proprietary instrumentation. The platform discovers Kubernetes infrastructure topology through OpenTelemetry Events, mapping relationships between pods, nodes, and services without requiring specialized agents. Organizations already invested in OpenTelemetry instrumentation can leverage existing deployments rather than replacing them with vendor-specific collectors. The platform supports both Prometheus and OTLP histogram formats for accurate metric distributions. Cloud platform integration works through standard APIs, ingesting CloudFront performance data from AWS and Stackdriver metrics from GCP, ensuring compatibility as cloud environments evolve. GeoIP enrichment adds location context to request patterns for geographic analysis. This standards-based approach keeps telemetry pipelines portable. Organizations switching observability platforms or running multi-vendor environments can maintain consistent instrumentation without vendor dependencies dictating their monitoring architecture. Unified Observability Without Compromise Kloudfuse 3.5 solves the fragmentation challenge enterprises face with AI adoption: natural language observability through MCP, FIPS-validated security for regulated sectors, granular cost controls, and standards-based instrumentation, unified in a single platform rather than across disconnected tools. The Self-SaaS architecture deploys within customer VPCs, ensuring data never leaves customer infrastructure while delivering 60–80% cost savings compared to traditional vendors. Trusted by Zscaler, GE Healthcare, Tata 1mg, and Automation Anywhere, Kloudfuse processes millions of events per second across over 700 integrations. Kloudfuse 3.5 proves that comprehensive observability, from conversational queries to federal compliance to predictable costs, doesn't require fragmented platforms, vendor lock-in, or choosing between capability and control. Learn more: • Learn more about Kloudfuse 3.5 capabilities • Sign up for a demo • Read the technical documentation Kloudfuse will demonstrate Kloudfuse 3.5 capabilities at the Gartner IT Infrastructure, Operations & Cloud Strategies Conference in Las Vegas, Nevada, from December 9–11, 2025. Visit booth 646 to see the Model Context Protocol Server in action. About Kloudfuse Kloudfuse is a unified observability platform integrating with over 700 diverse infrastructures, cloud services, and applications. By harnessing open standards like OpenTelemetry and Prometheus, Kloudfuse eliminates vendor lock-in while providing advanced capabilities across metrics, logs, traces, events, and real user monitoring. Deployed within customer VPCs, Kloudfuse ensures scalability, cost efficiency, and enterprise security. Trusted by leading organizations like Zscaler, GE Healthcare, Tata 1mg, and Innovacer, Kloudfuse delivers observability that enterprises can operate with confidence. Learn more at www.kloudfuse.com or follow us on LinkedIn. (Disclaimer: The above press release comes to you under an arrangement with NRDPL and PTI takes no editorial responsibility for the same.). PTI