CIO-Reinvented Blog - By Prashanth Rai

Journey from Decision Support to Decision Automation - 2

Continuing from an earlier post on the Journey of BI systems from being just decision support to decision automation. On this topic I read a couple of interesting articles or posts , first was a blog post by Tom Davenport the author of Competing with Analytics and an article titled "What Good are Experts" in the document "Breakthrough ideas for 2008" by HBR.

As discussed in the first post Data/information from BI systems are supposed to aid/assist management/information worker make decisions. But is this what is really happening?

As indicated by Tom Davenport

When my former Accenture colleague Jeanne Harris and I surveyed companies in 2002, and again in 2006, about their enterprise systems, “better decision-making” was the objective most frequently mentioned as the reason for implementing the systems. Of course, that’s an easy response to give, since virtually no company measures the quality of decisions. Yet in interviews with some of the companies, we found not a single effort to actually connect enterprise information with decisions.

As a solution to this he suggests the following before any BI project is embarked in an organization

Projects need to start with an inventory of key decisions. For each one, a company could note who’s responsible for making it, how often it gets made, and what information and knowledge are necessary to make it well. A second approach would be to precede any effort to build business intelligence solutions—the type of IT that is most closely related to decision-making—with identification of the key decisions that would be made on the data and analysis. A third would be to begin assessing managers not only on the outcomes of their decisions, but on the processes they employed.

While on this topic lets take a deep dive into types of decisions. In the article "What Good are Experts" problems are categorized into two types

1. Rule-Based
2. Probabilistic

And suggests the optimum solution approach for these types of problem as highlighted in this graphic  

The Business Intelligence solutions of the future needs to be able to address all these 3 solution types, that is should be able to support rule based engines for computer driven decision making, should be able to support internal prediction markets/model which leverage the Wisdom of the Crows and finally be able to provide the experts all the required information so that they can better leverage their expertise and provide the best solutions.

Tags: business+intelligence - BI - decision - DSS - Hyperion - Oracle - SAP - Tom+Davenport - Competing+Analytics

Prashanth Rai

Journey from Decision Support to Decision Automation - 1

BI Applications sit a top the information value chain that originates in operational systems such as Core Systems, and other transaction processing systems, providing insights into issues such as customer profitability, corporate performance, risk and anything that is based on the collection and analysis of large amounts of data.

Though this is a critical tool from an organization / managements perspective, till very recently the power of this tool was left in the hands of only a few specialists. As the process of accessing the information in this system was quasi technical and needed more knowledge/training then an end user to be able to access the information. These specialists would act as information brokers and field all requests raised by the business and be responsible for generating the required information. This information would then need to be interpreted "manually".

In this process there were two types of information requests the regular and the adhoc. The regular reports which are institutionalized - they are fixed data set and fixed points of view and they are communicated to the team. This is the low hanging fruit of getting a BI system in place.

But the true value comes in the adhoc or on need analysis, When deep diving into a certain issue / problem the information worker or knowledge worker needs the relevant information / data that can assist him in getting insights into the problem at hand and also help in framing a solution for the same. But these kinds of requirements are not catered to sufficiently currently. In most cases when such requests go to the IT departments the elapsed time taken to be able to create information chain is so long that the time context is lost on the same.

Primarily in BI systems there are two models, One which sits directly on top of the ERP or transaction processing system and accessing the same in real time to serve the information requests. The second model is where this information is extracted transformed and loaded into Data warehouses and those warehouses then power information cubes/ OLAP engines which then serve the information worker.

Both these models have their relevant pros & cons the data warehouse based model - provides consistent, reliable data plus the benefits of database server technology designed to handle large sets of summary and detail-level data. However they take a long time to build and as far as BI systems are concerned, the value they provide diminishes rapidly the longer you take to deliver them. A warehouse project might get signed off based on some immediate market opportunities your company has spotted, but if you take a year to deliver the system, by then the opportunity may well have gone.

In the other model that directly sits atop the transaction system , they work by mapping their metadata layer directly against the underlying source data, applying aggregates and caches to try and speed up query performance. In some cases, these tools actually allow you to integrate data across multiple systems, though this will never be as fast to query as data that actually resides in just one database. Over time, systems built like this get harder and harder to manage, but at least you get reports and analysis in people’s hands fast and the business aren’t usually interested in arcane technical discussions about the merits of a properly architected data warehouse.

As indicated in the Rittman Mead blog, The possible solution to this is a hybrid between both these routes where the systems that directly sit atop the transaction system , are used for the immediate data access and over a period of time they be moved into a data warehouse and the BI application run from that as a data source. - Sounds quite promising.

But given this background it is important not to restrict our thinking at this point, what does the future hold for BI environments. "Decision Automation"

Decision Automation has been the holy grail of Enterprise Software for a very long time now. So are we any closer to this vision today ? (To be continued)

Tags: business+intelligence - DSS - bi - information - oracle - hyperion - sap

Image Source: 1

Prashanth Rai

TCS - New Organization Structure

Starting April 1st TCS will be adopting a New Organization structure, TCS is the largest Indian / Offshore IT Services provider with close to 100000 on its roles.

"As we scale up over 1,00,000 employees, TCS needs a structure that allows us to build a nimble organization to capture new growth opportunities," Mr S Ramadorai, CEO and Managing Director, TCS said. The organization has been restructured into five units, which will provide customers a single view of the company. The units would address new growth opportunities, provide sharper accountability and groom new leadership, its COO and Executive Director N. Chandrasekaran told.

N Chandrasekaran, chief operating officer, TCS, said: “It (the earlier structure) was great five years back when we were only 15,000-20,000-people strong. Today, with over 1,00,000 people across multiple geographies, there is a need for a new structure.”

The new structure, defines the organization into five ISUs — industry solutions group (ISG), major markets group (MMG), new growth market group (NMG), strategic initiative group (SIG) and organization infrastructure group (OIG) - would be headed by a director, reporting directly to the COO.

The Five ISU's explained:

The ISG unit will have multiple units dedicated to all the verticals that the company has presence in. All the existing clients will come under this group. Each unit under this will be of manageable size (which the company describes as having 2,500-5,000 people). Since BFSI segment is the largest segment with more than 35,000 people. This would be further divided into four units for better focus and manageability. The director of each these units will have complete empowerment on developing the resources, technical expertise, HR support, and financial support among other needs.

The major markets' unit will focus on geographies of the US, UK and Europe. The unit will be responsible for getting prospective and new client wins. While the group is focused on geographies they will also be responsible for brand building, maintaining relations with the analysts' community and the customers.

The new growth markets group will have three different units focusing on- India, APAC and emerging markets. These three units will own end-to-end responsibility from client wins to maintaining long-term relations.

The fourth group- strategic initiative unit will have three important units. First, is the TCS financial solutions unit, which was announced in the first quarter of this financial year. Second, soon to be announced small and medium enterprises business solutions and finally, a new unit for platform BPO solutions.

Finally, the organization infrastructure group will have four units, which will look into issue like process excellence, technology excellence, shared services group and resource management group. Each of these will again have individuals of proven capabilities and expertise.

Tags: TCS - Offshore - Outsource - India - Org+Chart - Future - CIO - Technology

Source: 1, 2, 3, 4, 5

Image Source: 1

Prashanth Rai