• Determine Business Challenge, Use Case Definition and Success Criteria

The first goal is to identify one initial business challenge and use case that focuses the efforts of all parties involved to build an end-to-end cloud analytics solution that delivers business value from the start. Deciding on a use case will determine which data sets to extract from which operational data sources, which edge-analytics to use, at what frequency, the metadata needed for enrichment, the target Amazon Service, the ML and analytics models and the visualization of results.

• Implement Security and Data Governance for Operational Data Source

On the operational IT side, Twin Talk is integrated with the chosen operational data sources and configured to match cyber security requirements and processes including VPC, firewalls, IPsec tunnels, etc. to be able to extract data and move it into the Amazon cloud. Data governance is implemented via Twin Talk’s Asset Catalog Manager. Data pipelines are set up within Twin Talk to enable the transport of large amount of historical data as well as continuous delivery of real-time data.

• Implementing Cloud Architecture via Templates and Blueprints

Meanwhile, TEKsystems Global Services is building out the event-driven production and monitoring cloud architecture by leveraging proven expertise, blueprints, methods, and tools which dramatically accelerates time-to-value and reduces the resource commitment.

• Implement Analytics, ML, Applications and Visualization

Once the OT and IT components are in place and end-to-end data delivery is working, Twin Talk will transport historical data into the Amazon Cloud services so that analytics and ML tools and models can be configured, trained and tested. Next is the development of visualization and dashboards that can display anomaly detection results and alerts.

• Event-Driven, Real-time Operations and Business Value Delivery

Lastly Twin Talk is used to extract, enrich and transport real-time operational data into the Amazon Cloud services that are leveraging the previously trained analytics and AI models, visualizations and dashboards to detect opportunities for production.