Digital Twin

  • 29th, Nov 2020

1. AI Systems are needed for DevOps to Cloud migration An enterprise cloud strategy requires a serious, operational data strategy starting with AI/machine learning that can mature the enterprise migration capability. Often noted is that AI does save IT a ton of tedious manual legwork and the costs and headaches that go with it. But more importantly, AI with systemic automation helps to lay the foundation for a more mature culture where developers, operators and reliability engineers can live and breathe DevOps together. A fully-formed, agile DevOps culture, facilitated by AI and automation, is the key to a successful cloud transformation journey. Successful Cloud migrations & productions rely on:

 Automation of continuous builds, integration and delivery;

 Automation of operations, performance monitoring and instrumentation for monitoring;

 Automation of root-cause analysis and remediation;

 Automation of performance baselining and configurations;

 Accountability of CI/CD pipelines in meeting deadlines and delivering software reliability;

 Software intelligence needed to close existing automation gaps at decision gates and the Use of validation and performance signatures to execute testing for new builds & production scenarios.

2. AI Profiling and Operational Automation is needed for Legacy Enterprise Applications It is imperative to know how legacy applications work. Baselining their performance to compare how they’re performing in the cloud later is an obvious first step. But it’s so tedious that it begs for portable and extensible automation as a necessity. AI is built to baseline and predict how systems and apps are performing today and how they need to perform in the cloud. Monitoring plays a key role in this profiling stage: starting from creating topology maps of the entire technology stack; mapping out interdependencies between systems; and automating performance baselining, through to full stress tests. AI with ML can automate these operational steps to ensure a comprehensive snapshot of the existing system architecture, service flow, and performance are carried forward to the cloud.

3. AI Systems provides verifiable insights to Operate Cloud Production A combination of AI-derived real-time insights and user-session analysis for cloud migration enables automating performance monitoring, remediation, CI/CD pipelines, root-cause analysis, testing, system configurations, and many more steps. This empowers software engineers to have full ownership over the entire value chain: from initial coding to deployment and reliability of the final product. That level of visibility gives its DevOps team instant feedback, which in turn informs and expedites the pace of software innovation delivery. Software intelligence builds on strong AI to oversee the health of the entire system from end to end. Smart anomaly detection, automated testing, real-time visibility, and business impact assessment are indispensable key pillars of support that AI brings to the table. 

4. AI/ML workloads address Data and Cyber Security concerns for Cloud Migration In the past, enterprise security was primarily based upon securing a network perimeter around the enterprise data center under the premise that all systems, data, and resources within the perimeter were secure and could be trusted. Moving to Cloud and using the vast amount of data for analytics specifically for any AI/ML initiative magnifies data and cybersecurity concerns. For this kind of usage, we need vast amounts of data and also the data needs to be un-masked to facilitate model exploration, training, and verification activities. Data visibility and explore-ability requirements in the cloud demand new disruptive approaches to address cybersecurity and data breach threats. Today the foundation of secure enterprise cloud tenancy, through the benefit of AI software agents, needs to:

 Provide Identity based security;

 Establish Role-based Access Controls, and

 Develop and Implement Zero -Approaches and Zero-Leakage policies.

However, entropy eventually causes the cloud configuration to change or drift in unexpected ways. Interestingly, AI-based monitoring of the enterprise’s cloud tenancy can anticipate and flag any unexpected configuration changes made by IT, or those introduced by the cloud vendor, ensuring that any cybersecurity vulnerabilities are predicted, caught, and rectified before they cause significant damage.

5. AI-Driven Cloud Service Broker is purpose-fit for cloud production in a Hybrid Cloud Universe Enterprises are now envisioning a bimodal IT strategy that increases the reliability of its current infrastructure while creating a more flexible infrastructure that allows for experimentation and innovation to support mission needs. A bimodal approach drives change by allowing to optimize existing IT infrastructure (Mode 1) while pursuing exploratory methods for solving current enterprise challenges and implementing rapid, strategic, and innovative solutions for the business (Mode 2). Mode 1 is characterized by the focused application of current technologies to critical targeted areas such as data center facilities and operations. Mode 2 is characterized by exploration and innovation, which emphasizes IT agility and speed in support of business operations. This next-generation IT vision drives the enterprises to put in place a hybrid cloud infrastructure that combines existing on-premises private infrastructure with public and community cloud infrastructures. The intent is to rationalize applications and workloads across the secured hybrid infrastructure, allowing portability where necessary between on-premises and Cloud Service Providers (CSP). AI and automation can manage the complexity of hybrid cloud architecture and environment, as well as provide access to hyper-scaling utilizing multiple CSPs. As an example, AI can drive the portability for mission-critical applications, where automation steers the capabilities such as abstraction and container technologies (like Kubernetes) to allow mission-critical applications to be multi-CSP, and run on multiple CSP at once or to very quickly failover from one CSP to another with no data loss.

The AI Resilience Lab™ is an integral part of the FS.AI AI Scaffolding™ methodology. This abstract is one, in a series, produced to summarize advantages recently demonstrated in that setting. If you are interested in more information on this topic, the AI Lab process, and/or Scaffolding method please contact: info@finservai.com and go to www.finservai.com and follow us on www.linkedin.com/company/finservai

Download Version