Ops, AI Did it Again

Publish date:

AI comes to the rescue of complex IT operations, improving step-by-step efficiency and reliability while it learns, on its way to full autonomy

So many systems, services, devices and applications swarming around in an enterprise IT operations landscape. So much data available in real-time about how they perform, succeed and fail. It’s the perfect playground for AI to get a grip on the complexity, by learning from IT operations data to provide improvement. First by giving better insight into the performance of operations and by real-time detection of disturbances. Then – through predictive analytics – by anticipating these disturbances, so that timely measures can be taken. Finally – when it has found even the most complex, hidden patterns – by autonomously optimizing IT operations. Oops, is that infrastructure simply taking care of itself?

What

  • AI for IT Operations (AIOps) – also sometimes referred to as Algorithmic IT Operations – aims to collect and analyze IT operations data, often in real-time, in order to continuously fix and improve IT operations’ performance.
  • Data to drive AIOps can be ingested from multiple and diverse sources including; logfiles, IT operations management platforms, problem ticket data, connected devices, ‘wire’ network traffic data and event monitoring / alert systems.
  • In a way, AIOps may be considered as applying and extending the principles of continuous integration and delivery (CI/CD), and DevOps to the core functions of IT operations.
  • AI / machine learning and (intelligent) process automation are key approaches within AIOps, to explain and understand the current situation, predict what may happen, prescribe what needs to be done to achieve performance objectives, and eventually, have IT operations run and optimize itself in (semi) autonomous ways.

Use

  • A large global car manufacturer is using an agile and AI infused DevOps capability to accelerate one of its own key IT capabilities.
  • A large European-based Postal Service applied DevOps based capabilities across all its mission critical services, reducing outages by over 50% and increasing speed by twenty times.
  • Brazilian telecommunications provider, Nextel provides uses AIOps to monitor more than 25-thousandnetwork elements, reducing the time to respond to network incidents from 30 minutes to less than 5 minutes.
  • Cisco achieved 11 million dollars in annual savings on its incident resolution activities, through applying AIOps and automation.
  • Vodafone correlates data across all its infrastructure layers – from critical business and application services to underlying IT and network components – to discover the root cause of problems and their related business impact up to 3 times faster.
  • cyBERT is an ongoing experiment to train and optimize transformer networks, to flexibly and robustly scrutinize very large cybersecurity log files. Using AI-driven natural language processing it effectively analyzes the log files, looking for anomalies, trends and other insights.

Impact

  • As discussed in the Capgemini 2018 study, ‘The automation advantage,automation can deliver:
    • 76% improved company profitability
    • 87% faster product and service delivery
    • 82% improved software development
  • Clearly, many of the generic advantages of automation also pertain to IT operations, when viewed as ‘just another’ business activity area.
  • Automate and augment routine, repeatable IT operations tasks, so that staff can focus on more strategic, value-adding activities.
  • Without automation, tasks would need manual execution, increasing complexity and risk of error, which in turn can lead to outages.
  • Automation and AI / machine learning tools pave the way to a full self-service infrastructure and applications management landscape.
  • Infrastructure and applications management trends that may result in outages are detected and mitigated before their impact is felt.
  • Better delivery of SLAs and increased customer satisfaction, e.g. through faster problem resolution and less outages.

Tech

Empfohlener Experte

Marc Bauer

Expert in Strategy and Transformation

Related Resources

Reinventing IT – Podcast

Inventive IT ist eine digitale Realität in der Erfindungsreichtum erforderlich ist. Dieser...

The Great Digital Divide – Podcast

Die COVID-19-Pandemie verändert weltweit wie wir leben, arbeiten und mit anderen Menschen...