Technology Architect – Data & AI & Co-Founder at PinCarts Private Limited Sep 2016 - Present

PinCarts is technology enabled platform with a vision to provide a seamless shopping experience to its valuable customers throughout their shopping journey using futuristic retail technologies like proximity marketing, personalization, advanced retail analytics, indoor navigation, contact-less checkout, and virtual and augmented reality. PinCarts’ business model for brand retailers and shopping malls is meant to benefit the shoppers and retailers both. This technology helps with comprehensive cloud-based pre-sales and post-sales analytics and facilitates key decision-making.

Core Responsibilities:

• Strategic planning for the company's future growth, including technology and business operations, marketing, and sales strategy development.
• AWS Cloud Infrastructure set-up for various Data, Analytics, AI, Web, and Mobile applications.
• Built an end-to-end Cloud Data Lake and Analytics System with data from varied data sources (structured/unstructured) using various technologies like Cloudera Hadoop Cluster, Apache Spark, Kafka, Impala & StreamSets ETL, SQL & NoSQL MongoDB, and AWS S3.
• Designed an end-to-end analytical Cloud Data Warehouse and BI solution with AWS Redshift Data Warehouse architecture with ETL/ELT services for Data Ingestion pipelines and Data Sharing with Data Analytics Services and Apps for BI and Machine Learning applications.
• Cloud Infrastructure set-up and Deployment of Web & Mobile Apps with AWS EC2, Route 53 DNS, Load Balancer, CloudFront, HTTP Web Server, Apache Tomcat app server, and MySQL.
• Designed and implemented a Content-Based Fashion Recommendation System using Hadoop HDFS, Hive/Impala, Apache Spark, Mahout, Java, Python, and Elasticsearch.
• Business-to-business Sales - On-boarded over 25 retail customers with end-to-end system integration for product inventories to provide digital and in-store navigation of products.
• Proximity marketing platform using various location technologies like IoT, BLE Beacons (Eddystone/iBeacon) and geofencing for tracking customers in real-time.
• Setting up the GPU Deep Learning infrastructure with NVIDIA CUDA Toolkit with Python TensorFlow/Keras on Linux platform for building end-to-end machine learning pipelines.

Launch Press Coverage:

Business Standard    Yahoo News India

Technical Architect at Hexaware Technologies Inc. May 2006-Jul 2016

Designation: Technical Architect
Duration: May 2006-Jul 2016

Clients:

• Northern Trust Corporation
• American International Group (AIG)
• Tustmark Insurance

Below are the project details.

Global Accounting Hub and Operational Data Store May 2015-Jun 2016

Client: Northern Trust Corporation
Role: Technical Architect
Work Location: Chicago, IL

Global Accounting Hub is one of the strategic accounting projects under the Asset Servicing program. The goal of strategic accounting is to consolidate the posting of transactions from the various systems in use today into the new accounting system.

Enterprise data warehouse/operational data store is a Hadoop-based storage and processing management system. The purpose of this project is to isolate all the source systems from reporting requirements.

• Implemented operational data store for accounting data in Enterprise Data Warehouse using Hadoop and Apache Impala/Hive metastore framework.
• Designed and implemented strategy for providing accounting data required for Wealth management system processing.
• Designed overall data transformation strategy for publishing the real-time transactions into message queue using IBM CDC and DataStage.
• Designed and Implemented service-oriented architecture for accounting business units.
• Integrated with upstream and downstream systems/applications with web services interface.
• This helped remove inter-system dependencies by defining ISO messaging standards.

Environment:
Apache Hadoop, Apache Impala, Java, IBM CDC, IBM InfoSphere DataStage 11.3, UNIX, Oracle Service Oriented Architecture (SOA) suite, IBM Mainframes and DB2

Comprehensive Capital Analysis and Review (CCAR) Oct 2014-Mar 2016

Client: Northern Trust Corporation
Role: Technical Architect
Work Location: Chicago, IL

• Designed overall automation strategy for production execution of all CCAR schedules, which ensures fully automated execution and rerun ability and email notifications.
• Implemented Business-As-Usual (BAU) ad-hoc data load process with fully automated execution mechanism using Unix shell scripts and DataStage dsjob utility.
• Designed and implemented Retail-Other schedule with 99% data accuracy in system integration testing with IBM InfoSphere DataStage 8.5 and Oracle database in Unix environment.
• Reduced time spent on quality assurance and decreased product defects by enhancing development procedures.

Environment:
IBM InfoSphere DataStage 8.5, IBM InfoSphere Server Manager and ISTOOL, WinSCP, UNIX, Oracle 11g

Policy Administration System Jul 2014-Sep 2014

Client: Tustmark Insurance
Role: Data Modeler
Work Location: Lake Forest, IL

• Ensued database guidelines to create a new relational database that will make all the data from the source data spreadsheets available to end users.
• Standardized data storage and normalized the table/entity relationships up to a certain extent to eliminate data redundancy and manipulation anomalies which helps automation test data users to analyze the input test data easily with simple SQL queries.
• Documented all business requirements (BRD) for logical data model preparation and future references.
• Prepared physical data model and DDL scripts for object creation.

Environment:
Microsoft SQL Server 2008, Visio Professional 2013

Global Data Warehouse (GDW) Jan 2012-Mar 2014

Client: American International Group (AIG)
Role: Team Lead
Location: Mumbai/Pune, India

This project was for implementing a global data warehouse system for the human resource department under Global Workforce Reporting & Analytics (GWR&A) initiative. GDW gets data from 4 source systems including domestic and international feeds and applies calculations to enable fact-based, consistent, and transparent analytical and operational reporting and other ad-hoc packages in order to facilitate decision-making.

• Delivered a First-of-a-kind multi-dimensional data warehouse for HR analytics and operational reporting with IBM InfoSphere DataStage and Oracle as a back-end database in Unix environment.
• Automated and scheduled data load processes using Unix shell scripts and PeopleSoft scheduler with interactive email notifications and error log distributions.
• Improved Data Quality and standardized definitions for metrics and data elements using common rules engine using DataStage routines.
• Streamlined data processing for downstream applications and security using the global security Model.

Environment:
IBM InfoSphere DataStage 8.5, PeopleSoft EPM 8.9, UNIX, Oracle 11g, PeopleSoft Scheduler

Northern Financial Objective Reporting, and Management Project (nForm) Dec 2007-Dec 2011

Client: Northern Trust Corporation
Role: Development Team Lead
Work Location: Chicago, IL

• Re-engineered profitability and trust asset reporting, costing, planning, and forecasting to drive better decisions, accountability, and profit by applying calculations to derive organization, client and product profitability.
• Spearheaded a development team of over 15 members disseminated across onsite and offshore locations.
• Modernized Northern's financial technology platform and updates methodologies to enable fact-based, consistent and transparent reporting and planning of profitability and other financial metrics in order to facilitate decision-making. Spearheaded a development team of over 15 members disseminated across onsite and offshore locations.
• Delivered organization and client profitability data marts with performance efficient design which ensured on-time delivery of financial reports to senior management. Designed and implemented multi-dimensional security for operational reports to ensure highly secured information governance.
• Automated and streamlined execution notification processes for uninterrupted schedule flow Using technologies like IBM InfoSphere DataStage Universe/XMETA Repository, UNIX, XML, XSLT, Java, Java XML Parsers (DOM and JAX), Apache Web Server, Java & VB Scripting, Oracle PL/SQL, Microsoft DOS Batch programming and FTP techniques.
• Modules/Data Marts:
  Financial Reporting, Trust Asset Reporting, Northern General Ledger Warehouse Reporting, Northern Trust Client Profitability, Cognos Planning and Forecasting

Environment:
PeopleSoft EPM 8.9, IBM InfoSphere DataStage, People Tools 8.4, Windows XP 2002, UNIX, Oracle 10g, Control-m, Rapid SQL