The Center for Data Innovation spoke with Maitreya Natu, chief scientist of Digitate, a California-based tech company that uses AI to automate IT and business operations for a wide range of companies. Natu discussed Digitate’s flagship product, ignio, and how future advances in generative AI will transform enterprise IT operations.
Martin Makaryan: What services does Digitate offer?
Maitreya Natu: At Digitate, our mission is to redefine IT operations through AI and automation. Digital transformation of enterprises has accelerated in recent years, and we want to create intelligent, adaptive systems that make enterprise IT environments more reliable and resilient. We provide several solutions such as AI monitoring of retail and e-commerce activities, event management, AI-based analytics, but our signature product is ignio.AIOps, an AI-powered IT operations analytics tool that helps companies make data-driven decisions when it comes to IT operations.
We work with a wide range of customers, from mid-sized enterprises to large corporations across various industries, including finance, healthcare, retail, manufacturing, and telecommunications. Our clients usually seek to optimize their IT operations, reduce operational costs, and enhance service reliability. They turn to us for our expertise in managing complex IT environments and our ability to support their digital transformation initiatives with AI-driven solutions.
Makaryan: How do you use AI?
Natu: AI innovation is the cornerstone of our offerings at Digitate. Our flagship product, ignio, leverages advanced AI to analyze vast datasets, predict IT issues, and automate routine tasks. This transformation allows our customers to achieve operational efficiency, reduce manual interventions, and ensure that their IT infrastructure can support the evolving demands of today’s business environment. For example, ignio helps retail stores achieve uninterrupted store logistics operations, enhanced end-user experience, and reduced incidents to avert business loss by providing predictive analytics on potential at different times. Indeed, our products can help businesses counter up to 40 percent of their IT issues before they occur. We also have an add-on product, ignio Observe, that uses machine learning algorithms to derive actionable insights for businesses by mining distinct patterns from log files of historical data.
Our philosophy is to transform IT workload management from a reactive to a proactive approach leveraging AI. There are a variety of AI features working at the back end to make this work. We first use process mining algorithms to understand the context of workflows in an organization. We then use prediction models built on historical data to predict the execution of these process flows. Then anomaly detection algorithms kick in to detect deviations from normal behavior and other models to assess the impact of failures and anomalies in the form of potential delays, for example. Thanks to this proactive approach, the IT teams do not spend their time dealing with surprise “fires”. The goal is to have predictable and agile operations.
Makaryan: Can you describe the AI models you have built?
Natu: We can classify the AI and machine learning models we use based on the type of reasoning: descriptive, diagnostic, predictive, and prescriptive reasoning. Descriptive reasoning answers the “what is happening?” question using various tools to create context of an enterprise system. These include natural language processing, transformers, process flow mining, pattern mining, etc. Diagnostic reasoning answers the “why?” question using root-cause analysis tools such as anomaly detection, Bayesian reasoning (an inference method for statistical reasoning), classification models, event correlation, etc. Predictive reasoning answers the “what is about to happen?” question and uses fault propagation models, deep neural networks, and time-series forecasting. Finally, prescriptive reasoning answers the “what is the best that can happen?” question.
Makaryan: What has been the biggest challenge for you as an innovator?
Natu: As chief scientist at Digitate, I lead the research and development of our AI and automation tools. One of the biggest challenges as a data innovator is creating applied AI solutions where theory meets practice. Real-world situations present several rubber-meets-the-road challenges such as poor data quality, very large data volumes, inconsistent user feedback, and a continuously evolving landscape, to name just a few. To push for AI adoption in the IT industry, it is essential to ensure that AI solutions are actionable, explainable, trustworthy, and accessible. Balancing all these competing considerations has been and remains a constant challenge.
Makaryan: How will future advances in generative AI shape Digitate’s products and services?
Natu: Generative AI will empower ignio to transform the human-in-the-loop experience in autonomous IT operations. Generative AI can transform the first part of the automation process that machines handle by creating automatic instructions for things like setting up resources, configuring services, and managing software updates. The second part, where humans step in, involves validating these actions, ensuring everything is correct, and fixing any mistakes the AI might have made. Generative AI can transform this stage of the process by creating an assistant to drive intelligent conversations. It leverages AI’s ability to understand language, capture user context, and learn from feedback. As a result, IT staff can consume the insights in a much simpler and intuitive way, leading to better incident resolutions and proactive problem management.
