| |February 20179CIOReviewCIOReviewPOD teams will typically comprise of the following components - combination of data scientists/statisticians, big data engineers, clinical staff, healthcare coding experts, application developers, visualization experts, tech writers and technology tools/enablers.The POD components in the framework will not be persisting as silos in their own departments but would work together under a POD owner who will create the right skill/effort capacity combination for the deliverable. The different components/elements will get support from a center of excellence within the vendor for any upskill/cross-skill training. The POD elements would collaborate in an intricate fashion through a well-designed framework and defined metadata for their roles to create the tapestry needed for delivery.The example of the POD model could be delivering analytics as insights for a Hospital chain on their costs of readmission & reasons for the same, and what they should do to reduce the costs. Typically this problem would be addressed by a product vendor with a BI reporting tool reading of data-marts, and queries configured in it by an analyst based on inputs from the Hospital side analysts. The second current method of delivery is to expand on the Hospital staff IT through an FTE model and this is the case when the Hospital has invested in a BI tool on its own but does not have the bandwidth to do the BI reports. The outcomes in both the above delivery models will be basic descriptive reports and these would be delivered with some difficulty and the timelines would not permit them to have a lot of insights. The growing government needs for more detailed reporting with insights will not be met with the current delivery model. The reason for the difficulties is not having the right combination of skills working in tandem as one small team to deliver value.Now, with the POD model, both the delivery models would be run seamlessly. The POD leader would understand the need of the Hospital and based on the problem statement, and assign a combination of resources ­ statisticians/data scientists with some primer knowledge of LACE model, reporting engineers with visualization experience for creating dashboards, clinical staff who understand readmission, coders who can read into the medical records and understand the costs ­ and plan capacity assigned to this assignment. This combination of resources with different skillsets along with tools travels through the project time as though they are a delivery POD. The initial small `POD' Structure will evolve into a full-fledged delivery model with multiple PODS working on multiple assignments from clients creating the POD-COE for Business Intelligence. The POD model is applicable for BI and Insights derivation, but not very useful for BPO type of work-products where repeatability is a premium and not very useful for IT services based work products as well. The BI product based vendors are stuck to their belief of plug/play in BI and it is not going to scale, and the FTE based teams which staff augments client side teams are not adding true value except for cost-arbitrage. BI is Knowledge based and hence needs these innovative delivery models, and BI-POD to POD-COE could be one of the premium delivery models to provide the much needed value.
< Page 8 | Page 10 >