Aadhaar, an Indian government initiative aims to provide a unique identity for every citizen. This involves managing, authenticating and reporting staggering amounts of data that requires a robust and scalable information management platform.
Here’s how we helped the Indian government manage this complex and critical initiative successfully.
The challenge
The Unique Identification Authority of India (UIDAI), envisioned a dedicated12-digit number as a unique identity for every citizen of the country. The number would be stored in a centralized database and linked to the basic demographics and biometric information (photograph, ten fingerprints and iris), of every individual. UIDAI needed a partner to design a credible mechanism for identity establishment, authentication, storage and retrieval. The system needed to meet the following objectives:
- Scalability to withstand enormous data volumes and throughput needs
- Availability of information through Business Intelligence (BI) systems and portals to be used by diverse users such as the public, UIDAI and academicians
- Building a data warehouse that has zero knowledge of the resident, thereby ensuring security of data
- Flexibility to store data at the most atomic level, ensuring data aggregation at any level for reporting
- Usage of open source technologies and tools to avoid vendor lock-in
Our solution
Mindtree was chosen by UIDAI as the primary application development partner. We collaborated with UIDAI and devised a solution strategy that comprised the following:
- A data warehouse to capture all enrolment related information except PII (Personal Identifiable Information) data of over 1.2 billion residents of India
- Solution design to effectively manage 8 peta bytes of raw data
- An efficient data management system to handle over 100 million database transactions per day
- Asynchronous data acquisition from transactional systems
- Dataset generation and distribution platform enabling consumption by portals and dashboards, data marts and analysis tools
- Use of technologies that leverage data locality, which support distributed processing for dataset generation
Our approach
- Designed a layered architecture for flexibility and scalability of data collection and information dissemination
- Used big data open-source technologies such as Hadoop, Hive, Pig to handle the scale
- Developed MySQL-based Operational Data Store (ODS) to cater to operational reporting
- Asynchronous data acquisition from transactional systems using an event publish / subscribe framework
- The framework uses message queues to publish data to the data warehouse without impacting the enrolment process
- The framework is also used to publish data on authentications
Business impact
- Scalable architecture to handle ten billion enrolments
- Easy monitoring of the enrolment process through dashboards and customized reports
- Informed decision making through a ‘self-service’ portal with relevant metrics and custom reports
- Prevented potential fraud cases through formidable checks at the enrolment stage
