How Engineers Can Use AIOps to Innovate Their Infrastructure

Paul Constantinides
Salesforce

In today's fast-paced AI landscape, CIOs, IT leaders, and engineers are constantly challenged to manage increasingly complex and interconnected systems. The sheer scale and velocity of data generated by modern infrastructure can be overwhelming, making it difficult to maintain uptime, prevent outages, and create a seamless customer experience. This complexity is magnified by the industry's shift towards agentic AI.

The need for a new approach to IT operations is critical, one that moves beyond manual monitoring and static thresholds to intelligent, automated, and proactive systems. At Salesforce, we've embraced this challenge head-on by pioneering AI for IT operations (AIOps). We're already seeing 2,800 engineering hours now saved weekly on Warden AIOps, an AIOps agentic platform to help our site reliability engineers (SREs) and service owners proactively detect, diagnose, and remediate issues faster with minimal manual effort.

This isn't just about managing scale; it's about building an intelligent, proactive, and fully autonomous system that frees our engineers to focus on keeping services up and running smoothly, not constant firefighting.

The Challenge: From Manual Monitoring to Intelligent Automation

Managing vast and intricate systems involves a significant amount of manual effort. Our SREs and service owners often found themselves "glass watching" — staring at dashboards across disparate systems to identify issues. This reactive approach, while necessary, was inherently limited by human capacity and the sheer volume of data.

This challenge led to the creation of Warden AIOps, our system that leverages AI to assist with operational tasks. Our vision for Warden AIOps is to transform day-two operations, the ongoing management, maintenance and monitoring of a system after its deployment, by moving from manual, reactive interventions to automated, proactive, and safe operations. In doing so, we've built a system that can take actions like automatically adjusting resources, restarting pods, or running custom scripts to safely prevent outages before they happen.

A New Era of Proactive Operations

Here's how Warden AIOps is helping our engineers with quick and automated resolution, improving overall service availability:

  • Intelligent Anomaly Detection with Merlion: One of our foundational breakthroughs was the development of Merlion, an open-source library that we developed specifically for the purpose of anomaly detection. Merlion combines traditional models like isolation forests and statistical models with sequential neural network models. This allows us to identify subtle deviations and predict potential issues before they escalate into incidents. We also developed Moirai, an open-source foundation model for time series forecasting, which predicts potential spikes or dips in our systems.
  • Unified Observability for Comprehensive Context: To achieve truly intelligent operations, we needed a unified view of our vast and complex data. We aggregate three petabytes of data daily from various sources, including metrics from service level objectives (SLO) metrics, custom metrics, events, logs, and profiling and diagnostics. This eliminates the manual effort of sifting through different dashboards, allowing our systems to correlate information and give engineers a full contextual understanding.
  • From Correlation to Causation (and Remediation): Our PyRCA open-source library developed by the Salesforce Research team, helps us analyze hundreds of telemetry, dependency graph, and tracing data points to pinpoint root causes, significantly reducing the time for humans to identify key signals. We also use generative AI to auto-generate Root Cause Analysis (RCA) and Problem Review Board (PRB) reports and an orchestration engine to take immediate, rule-based actions to mitigate incidents, such as restarting app servers, even while the true causation is being investigated.
  • The Agentic Leap: Reasoning Like Humans, at Scale: Agentic AI adds a "reasoning layer" on top of our anomaly detection. Our system can now describe anomalies in natural language, correlate metrics, and reason like a human, using context to determine if a signal is truly anomalous. This capability automates log anomaly detection and allows engineers to dynamically explore problem patterns.

The Road Ahead: Towards a More Autonomous Agentic Enterprise Future

Our journey with AIOps is continuously evolving. The integration of tools like Cursor with Warden AIOps, via Model Context Protocol (MCP), is paving the way for a more autonomous state, a "flow state" where developers and service owners can easily transition from a signal to identifying the problematic code with repercussive context (even business impact), and taking necessary actions.

We are building an agentic enterprise future where our infrastructure is not just managed, but intelligently self-optimizing and self-healing. 

Warden AIOps is an internal Salesforce AIOps platform, and Merlion, Moirai, and PyRCA are open-source tools. These technologies are not available for sale.

Paul Constantinides is EVP of Engineering at Salesforce

Related Links

Hot Topics

The Latest

The Real Enterprise AI Risk Isn't Hallucinations

Ask most IT leaders about their biggest concern with AI and you'll hear the same answer: hallucinations ... Today, however, the conversation has shifted ... As organizations move beyond chatbots and experiments, they are increasingly deploying AI agents that perform multi-step tasks. These systems retrieve documents, query databases, call APIs, generate reports, write code, and make recommendations. The issue is not whether the model can reason. The issue is whether the organization can see, verify, and govern the decisions being made along the way ...

Beyond Buy vs. Build: The New Playbook for Data Pipelines

While organizations want to take control of their telemetry, building telemetry pipelines from scratch can be a very daunting, complicated task, even when leveraging open-source standards like OpenTelemetry. It requires specialized knowledge across distributed systems, data engineering, and security. This fragmented approach across systems causes higher operational costs; it puts a strain on resources and reduces efficiency as teams have to work with different interfaces and processes ...

The Corporate Network Is Fading - Here's What Replaces It

For decades, enterprise networks were designed around a simple assumption: work happened inside the office. Applications lived in centralized data centers, employees connected through internal infrastructure, and security focused on protecting the perimeter that surrounded everything ... But the way organizations operate today bears little resemblance to that environment. Cloud platforms host critical applications, employees connect from homes and airports as often as they do from offices, and partners collaborate through shared systems that exist far beyond corporate walls. In short, the corporate network no longer resembles the environment it was designed to protect ...

IT Leaders Must Establish a Network Data Architecture Practice

As an analyst who researches how IT organizations design, build, and operate their networks, I find that network data is a constant source of pain. Network teams struggle with data quality, fragmentation, authority, access, and trust. And these issues undermine everything they try to do. Here are the numbers: Only 45% of network teams are completely confident in the accuracy of their network source of truth, which documents the intent of their network ...

Data Center Outages Are Decreasing but More Costly

The 2026 Global Data Center Survey from Uptime Institute reveals an industry navigating workforce constraints, escalating outage expenses, even as rising costs remain the top concern for management teams ...

Observability Has a Physical Layer Now: What AI Infrastructure Teaches IT Operations

The next observability gap may not be in the code. It may be under the rack. That sounds strange until you think about how AI incidents actually feel in the middle of an investigation ... The application dashboard may be accurate. It may also be stopping at the wrong boundary. AI systems depend on software, but they also depend on a dense physical stack: racks, power paths, thermal margin, maintenance activity and, in many environments, liquid cooling. Those physical dependencies can change slowly before they look like a software incident ...

TLS Certificate Expiration Is Becoming an Observability Problem

Certificate expiration is the rare outage you can see coming. Every TLS certificate carries the date it stops working, so the moment it will begin breaking connections is knowable in advance. That's what makes an expired certificate such a frustrating way to lose a service. What's changing now is how often that date comes around ...

Enterprise Cloud Observability: Choosing the Right Starting Point

Enterprises operate different combinations of workloads across cloud, hybrid and multicloud environments. For business-critical workloads, teams need to consider monitoring and observability early so they can detect health issues, investigate failures, and understand operational impact. Organizations place workloads on cloud platforms based on a combination of technical requirements, economics, existing dependencies, organizational standards, and business priorities. Their monitoring priorities therefore depend on what they operate and where those systems run. Those priorities will not look the same for every organization ...

AI Is Making Incident Management More Complex - Bad Data Makes It Harder

Top-performing businesses prioritize data-driven decision making, enabling leaders to move from intuition and gut feel towards evidence-based judgment. But that judgment is only sound when the data underpinning decisions is accurate. With incident management, data accuracy is particularly important. Long-term revenue, customer trust, and operational stability depend on high-quality data that enables teams to quickly identify and address the root cause of major incidents. Against this backdrop, governance becomes a critical endeavor to ensure the right data drives the right action ...

MEAN TIME TO INSIGHT Podcast - Episode 26: Network Compliance

In MEAN TIME TO INSIGHT Episode 26, Shamus McGillicuddy, VP of Research, Network Infrastructure and Operations, at EMA discusses network compliance ... 

How Engineers Can Use AIOps to Innovate Their Infrastructure

Paul Constantinides
Salesforce

In today's fast-paced AI landscape, CIOs, IT leaders, and engineers are constantly challenged to manage increasingly complex and interconnected systems. The sheer scale and velocity of data generated by modern infrastructure can be overwhelming, making it difficult to maintain uptime, prevent outages, and create a seamless customer experience. This complexity is magnified by the industry's shift towards agentic AI.

The need for a new approach to IT operations is critical, one that moves beyond manual monitoring and static thresholds to intelligent, automated, and proactive systems. At Salesforce, we've embraced this challenge head-on by pioneering AI for IT operations (AIOps). We're already seeing 2,800 engineering hours now saved weekly on Warden AIOps, an AIOps agentic platform to help our site reliability engineers (SREs) and service owners proactively detect, diagnose, and remediate issues faster with minimal manual effort.

This isn't just about managing scale; it's about building an intelligent, proactive, and fully autonomous system that frees our engineers to focus on keeping services up and running smoothly, not constant firefighting.

The Challenge: From Manual Monitoring to Intelligent Automation

Managing vast and intricate systems involves a significant amount of manual effort. Our SREs and service owners often found themselves "glass watching" — staring at dashboards across disparate systems to identify issues. This reactive approach, while necessary, was inherently limited by human capacity and the sheer volume of data.

This challenge led to the creation of Warden AIOps, our system that leverages AI to assist with operational tasks. Our vision for Warden AIOps is to transform day-two operations, the ongoing management, maintenance and monitoring of a system after its deployment, by moving from manual, reactive interventions to automated, proactive, and safe operations. In doing so, we've built a system that can take actions like automatically adjusting resources, restarting pods, or running custom scripts to safely prevent outages before they happen.

A New Era of Proactive Operations

Here's how Warden AIOps is helping our engineers with quick and automated resolution, improving overall service availability:

  • Intelligent Anomaly Detection with Merlion: One of our foundational breakthroughs was the development of Merlion, an open-source library that we developed specifically for the purpose of anomaly detection. Merlion combines traditional models like isolation forests and statistical models with sequential neural network models. This allows us to identify subtle deviations and predict potential issues before they escalate into incidents. We also developed Moirai, an open-source foundation model for time series forecasting, which predicts potential spikes or dips in our systems.
  • Unified Observability for Comprehensive Context: To achieve truly intelligent operations, we needed a unified view of our vast and complex data. We aggregate three petabytes of data daily from various sources, including metrics from service level objectives (SLO) metrics, custom metrics, events, logs, and profiling and diagnostics. This eliminates the manual effort of sifting through different dashboards, allowing our systems to correlate information and give engineers a full contextual understanding.
  • From Correlation to Causation (and Remediation): Our PyRCA open-source library developed by the Salesforce Research team, helps us analyze hundreds of telemetry, dependency graph, and tracing data points to pinpoint root causes, significantly reducing the time for humans to identify key signals. We also use generative AI to auto-generate Root Cause Analysis (RCA) and Problem Review Board (PRB) reports and an orchestration engine to take immediate, rule-based actions to mitigate incidents, such as restarting app servers, even while the true causation is being investigated.
  • The Agentic Leap: Reasoning Like Humans, at Scale: Agentic AI adds a "reasoning layer" on top of our anomaly detection. Our system can now describe anomalies in natural language, correlate metrics, and reason like a human, using context to determine if a signal is truly anomalous. This capability automates log anomaly detection and allows engineers to dynamically explore problem patterns.

The Road Ahead: Towards a More Autonomous Agentic Enterprise Future

Our journey with AIOps is continuously evolving. The integration of tools like Cursor with Warden AIOps, via Model Context Protocol (MCP), is paving the way for a more autonomous state, a "flow state" where developers and service owners can easily transition from a signal to identifying the problematic code with repercussive context (even business impact), and taking necessary actions.

We are building an agentic enterprise future where our infrastructure is not just managed, but intelligently self-optimizing and self-healing. 

Warden AIOps is an internal Salesforce AIOps platform, and Merlion, Moirai, and PyRCA are open-source tools. These technologies are not available for sale.

Paul Constantinides is EVP of Engineering at Salesforce

Related Links

Hot Topics

The Latest

The Real Enterprise AI Risk Isn't Hallucinations

Ask most IT leaders about their biggest concern with AI and you'll hear the same answer: hallucinations ... Today, however, the conversation has shifted ... As organizations move beyond chatbots and experiments, they are increasingly deploying AI agents that perform multi-step tasks. These systems retrieve documents, query databases, call APIs, generate reports, write code, and make recommendations. The issue is not whether the model can reason. The issue is whether the organization can see, verify, and govern the decisions being made along the way ...

Beyond Buy vs. Build: The New Playbook for Data Pipelines

While organizations want to take control of their telemetry, building telemetry pipelines from scratch can be a very daunting, complicated task, even when leveraging open-source standards like OpenTelemetry. It requires specialized knowledge across distributed systems, data engineering, and security. This fragmented approach across systems causes higher operational costs; it puts a strain on resources and reduces efficiency as teams have to work with different interfaces and processes ...

The Corporate Network Is Fading - Here's What Replaces It

For decades, enterprise networks were designed around a simple assumption: work happened inside the office. Applications lived in centralized data centers, employees connected through internal infrastructure, and security focused on protecting the perimeter that surrounded everything ... But the way organizations operate today bears little resemblance to that environment. Cloud platforms host critical applications, employees connect from homes and airports as often as they do from offices, and partners collaborate through shared systems that exist far beyond corporate walls. In short, the corporate network no longer resembles the environment it was designed to protect ...

IT Leaders Must Establish a Network Data Architecture Practice

As an analyst who researches how IT organizations design, build, and operate their networks, I find that network data is a constant source of pain. Network teams struggle with data quality, fragmentation, authority, access, and trust. And these issues undermine everything they try to do. Here are the numbers: Only 45% of network teams are completely confident in the accuracy of their network source of truth, which documents the intent of their network ...

Data Center Outages Are Decreasing but More Costly

The 2026 Global Data Center Survey from Uptime Institute reveals an industry navigating workforce constraints, escalating outage expenses, even as rising costs remain the top concern for management teams ...

Observability Has a Physical Layer Now: What AI Infrastructure Teaches IT Operations

The next observability gap may not be in the code. It may be under the rack. That sounds strange until you think about how AI incidents actually feel in the middle of an investigation ... The application dashboard may be accurate. It may also be stopping at the wrong boundary. AI systems depend on software, but they also depend on a dense physical stack: racks, power paths, thermal margin, maintenance activity and, in many environments, liquid cooling. Those physical dependencies can change slowly before they look like a software incident ...

TLS Certificate Expiration Is Becoming an Observability Problem

Certificate expiration is the rare outage you can see coming. Every TLS certificate carries the date it stops working, so the moment it will begin breaking connections is knowable in advance. That's what makes an expired certificate such a frustrating way to lose a service. What's changing now is how often that date comes around ...

Enterprise Cloud Observability: Choosing the Right Starting Point

Enterprises operate different combinations of workloads across cloud, hybrid and multicloud environments. For business-critical workloads, teams need to consider monitoring and observability early so they can detect health issues, investigate failures, and understand operational impact. Organizations place workloads on cloud platforms based on a combination of technical requirements, economics, existing dependencies, organizational standards, and business priorities. Their monitoring priorities therefore depend on what they operate and where those systems run. Those priorities will not look the same for every organization ...

AI Is Making Incident Management More Complex - Bad Data Makes It Harder

Top-performing businesses prioritize data-driven decision making, enabling leaders to move from intuition and gut feel towards evidence-based judgment. But that judgment is only sound when the data underpinning decisions is accurate. With incident management, data accuracy is particularly important. Long-term revenue, customer trust, and operational stability depend on high-quality data that enables teams to quickly identify and address the root cause of major incidents. Against this backdrop, governance becomes a critical endeavor to ensure the right data drives the right action ...

MEAN TIME TO INSIGHT Podcast - Episode 26: Network Compliance

In MEAN TIME TO INSIGHT Episode 26, Shamus McGillicuddy, VP of Research, Network Infrastructure and Operations, at EMA discusses network compliance ...