How to architect intelligent automation with ITSM: A real-world example

Insurance companies aiming for a satisfying and prompt experience are turning to intelligent automation by leveraging robotic process automation (RPA), artificial intelligence (AI), and machine learning (ML) to replace repetitive manual work, reduce human error, and enable faster processing with less cost.

[ Download the automation architect's handbook. ]

Integrated systems and processes allow customers to have real-time seamless experiences, but the infrastructure must be set up to support rapid updates to software and systems and the flexibility to undertake large volumes of data.

In my previous article, I presented an overview of how insurance companies can architect intelligent workflow automation. This article goes a bit deeper by examining how an insurance company achieved the full benefits of digitalization at the infrastructure level. Although this is based on an actual insurance company case, this pattern can be applied in many industries struggling with the same situation.

4 problems with traditional processes

We identified four problems that could be improved through digitalization.

Infrastructure challenges

Traditional infrastructure involves more manual work than modern solutions. Over time, as enterprises grow, the number of environments increases. Resources pile up, legacy library dependencies start to accumulate, and mystery configurations prevent quickly providing reliable infrastructure while maintaining security, compliance, and efficiency.

Human error

Without proper processes and management tools, tracking problems can be exceedingly tricky. It's challenging to determine who is accountable for what and match the system log with the request. It's often hard to avoid silly mistakes and misconfiguration when doing things manually.

[ Learn the top considerations for building a production-ready AI/ML environment. ]

Configuration drift

The lack of monitoring, inconsistent patching, and unrecorded manual changes leads to inconsistencies across configuration items, such as servers and devices. This drift makes the service unpredictable. Any minor change, such as redeploying, patching, or even updating libraries, threatens to cause unexpected downtime. The more unique a configuration item becomes, the more time it can take to recover after an incident.

Limited scaling and slow delivery

IT often becomes a bottleneck to delivering services to customers. When IT cannot promptly provide developers with an environment, the lack of collaboration can cause silos and trust issues.

Furthermore, security and policy are of the highest priority, but it's a painstaking job to make sure all are applied correctly as the number of servers grows.

3 steps to operational efficiency

There are three main steps when you're trying to achieve operational efficiency:

• Monitored and managed workflow
• Infrastructure as Code
• Automate everything everywhere

[ Download the eBook: Modernize your IT with managed cloud services. ]

How IT service management (ITSM) works

IT service management (ITSM) establishes processes and practices to optimize the use of IT services. This approach gives a company centralized control and better monitoring of its financial costs.

Image

In the diagram above:

1. All infrastructure setup and configurations are stored as code using Ansible playbooks. These are stored in a Git repository for version control.
2. ITSM is used to manage and support policies and processes for IT services. Users can place their request, which triggers an automated workflow or playbook to start provisioning, updating, or deleting the target resource on-premises or in the cloud. The components are all connected via application programming interfaces (APIs).
3. Credentials and configurations are securely stored in the controller (although you can configure them to point elsewhere), which can be accessed when executing the playbooks. The automation does not end at the target. Follow-ups such as scanning, syncing inventory, starting another business process, adding monitoring pieces, and other configuration management can run concurrently or sequentially with role-based controls.
4. The targets to be created can be any resources on the cloud or in datacenters, such as a Kubernetes cluster, storage, virtual machine, or network stack. If a problem or failure occurs during execution, an incident will be created and synced back to ITSM for further tracking or reviewing. Additional components such as Smart Management are included as part of the initial installation to add an extra layer of automated self-healing, with incidents recorded and sent back to ITSM. Red Hat OpenShift running on the cloud helps manage and integrate these components.
5. The configuration management database and Ansible repository are in-sync with two-way communication. This can keep accurate records of assets across multidomain environments, disparate users, and teams so that all infrastructure information is up to date.

Automation and integration

In this setup, automated processes replace manual work with more compliant and effective operations while reducing the labor cost. It's consistent and can make dynamic adjustments as it enforces guardrails to deliver infrastructure services. This has increased insurance companies' agility, allowing them to scale to meet growth, and it can do the same for other industries.