State-wide Analytical Engine
Engagement Details
Primary Operations: India
Associated since: 2017
Service Type:
Application modernization
Service Model:
Managed service
Status: Completed
Methodology: Agile
Scope/Challenges/Key objectives
Business Case:
The solution framework required was a state-wide analytical engine which takes in data from various sources like government department databases, internet, sensors, machine logs, etc., transforms them, and presents them in an user friendly format. The solution must provide insights into government schemes and policies through relevant performance metrics and to improve collaboration among departments. The client needs a system of dashboards which would enable administrators to monitor and implement government programs effectively.
Challenges:
- Data silos – 24 distinct state departments.
- Data discrepancies and completeness.
- No insights into performance of departments.
- Zero to low utilization of data in most departments
Benefits:
- Cost reduction by power optimization.
- Increased ROI for the electricity department.
- Forecast solution for vaccination inventory optimization and track the efficient usage.
Solution, Technical Environment, Benefits
Solution / Project Engagement:
- Development of integrated BI & advanced analytics solution.
- Structured/unstructured data from databases/web/IoT – design data models, perform ETL, improve quality, aggregate, make available for down stream applications.
- Self service BI, statistical/predictive analytics solutions.
- Price forecasting for commodities.
- Sentiment analysis on government created policies.
- Recommendation engine based on statistical models for taxation.
- Recommendation engine based on statistical models for Govt. Loans & swapping.
- Market basket analysis on MNREGA policy to increase average wage.
- Cluster analysis to improve farming practices & crop yield.
- Correlation analysis for crop yield and nutrient recommendations.
Technical Environment:
- SAS EMiner, SAS VA, SAS EG, SAS data quality, SAS contextual analytics, SAS forecast studio, microsoft SQL server 2017, hortonworks data platform.
Enhanced role based access–insights, predictions, forecasts, KPI’s
Data utilization - 70%
Prediction accuracy - ~15%
Reduced cost of interest rate by 2% on investment loans
