Analytics works on data-intensive problems in varied functional and sector domains leveraging advanced tools (e.g., SAS, CPLEX, @RISK etc.) and techniques (e.g., cluster analysis, linear programming, Monte Carlo simulation, etc.) to derive insights from raw data.

In particular, Analytics provides expertise in:

• Managing large data sets
• Analyzing data using advanced techniques
• Building complex models
• Developing tools to solve business problems

	Data warehousing and data mining of patient healthcare records Problem: A European national healthcare provider had been involved in a very large transformational effort to reform the healthcare delivery system. Response: We helped the consulting team in diagnosing opportunities for improvement by analyzing historical patient records across various hospitals in the country. The key challenge was to warehouse and mine an extremely large and complex data set of 80 GB. We approached this situation by creating a robust data-warehousing platform on a high-end dedicated server with SAS. This platform has been successfully leveraged by consulting teams for over 20 engagements and has been instrumental in improving operational efficiency for ~200 hospitals.
	Applying simulation to optimize inventory across the supply chain Problem: A world glass major was facing huge losses due to excessive inventory across a very complex supply chain. Response: We helped the consulting team understand the dynamics of their supply chain to decide optimum inventory level and their exposure to demand shocks. The key challenge was to model the complicated distribution system and the vast supply chain. We developed a two-step model – a robust discrete event simulation model followed by a heuristic based optimization model. The system is currently being used in inventory management under dynamic demand conditions and has been key to improving overall supply chain performance.
	Developing a credit risk scorecard model Problem: A large and fast growing Asian bank was facing increased defaults from retail (credit card) and SME (small and medium enterprise) customers. Response: We helped the consulting team develop a state of the art risk scoring system to monitor and approve credit. The work involved application of sophisticated statistical modeling techniques like logistic regression and CHAID trees. The recommended systems have been implemented successfully and have resulted in substantial savings through decreased default rates.