Summary
Overview
Work History
Education
Skills
PATENTS
Timeline
Generic

Swaranjit Dua

San Francisco Bay area

Summary

Technology Leader, Software Engineer and Architect with wealth of experience in high-tech companies leading the architecture, innovation and development of enterprise and distributed software. Highly motivated, visionary, innovative and excels at analyzing complex problems and creative solutions. Multi-industry experience.

Overview

30
30
years of professional experience

Work History

Principal Engineer

Moody’s Corp.
08.2018 - Current
  • Lead architecture & development of different services to import and export different kind of data to an SAS platform, which are eventually stored in AWS S3 buckets.
  • Lead design for a common AWS S3 tenant bucket architecture that is used by all RMS product & services, covering issues such as, multi-tenancy, life-cycle management and security. These different services run as micro-services on Kubernetes platform.
  • Designed Rest API Import and partition process for different peril Event Data into S3.
  • Designed a generic model development kit (MDK) to run any natural catastrophe model into a cloud platform as a SAS application, using Spark, Scala, Parquet, AWS S3. Implemented and tested MDK and Wildfire peril features. Designed many run time optimization strategies on spark.
  • Designed & implemented process to transform parquet loss files from S3 into relation data base format using Spark, Scala. Implemented various spark features like group, join, aggregation etc.
  • Re-architected metadata catalog query service to scale for large data, using Postgres database.
  • Identified new ecommerce product growth opportunities and go-to-market strategies that leveraged core products.
  • Integrated RMS core product with RMS authentication engine.

Architect

Criteo
03.2015 - 07.2018
  • Led design and co-implemented Campaign Structure Tool that automates Google Adwords Product Listing Ads (PLA) partition structures and uploads on Adwords. Built with Java Map/Reduce and Hadoop infrastructure for scaling large number of customers and campaigns. Designed and implemented Multi target, feed filter based campaigns infrastructure.
  • Analyzed Criteo Search Conversion data model using Logistic Regression, Deep learning models including image embedding’s trained through CNN. Used Tensor flow and AWS.
  • Designed and Implemented regression job for an ingestion pipeline using Spark, Scala.
  • Led Architecture and implementation of Flexible Campaign Structure and Templates Transition Automation tool, enabled AB testing of Structural Campaign Segmentation for optimizing Products bidding performance. Patent committee recognized innovation and decided to keep as secret sauce. Built with Java, Adwords API.
  • Implemented partial Query Bidding tool that assigns filtered search query terms as “negative keywords” to different PLA campaigns to have different bids based on keywords performance. Is implemented on Hadoop Map/Reduce.
  • Designed and Implemented a date based large data workflow, using high performance complex SQL queries, on Amazon RedShift, used for creating summary reports on Tableau server.

Principal Engineer

Datapop Inc
04.2012 - 03.2015
  • Technical leader for Datapop flagship product – semantic and targeted based search advertising. Lead the product development from inception to release by collaborating with sales, product management and development. Architected and developed build platform, build tools and analytical tools for content optimization. Architected SOA based build platform, CAMEL, to configure and deploy different services used in semantic advertising. Architected and implemented Java Rules engine for generating targeting system for semantic advertising, feed transformation (Product Listing Ads), AB Testing framework, feed diff tool and Ad offer generation tool. Designed and implemented semantic dominant attribute analysis. Designed and implemented feed clustering using KMeans Machine Learning (mallet). Designed Spring/REST web service based system for run time creative optimization and testing. Wrote python script to move data from Redshift to MongoDB.

Senior Principal Engineer/Lead Architect

Oracle Corporation
07.2009 - 04.2012
  • Technical leader for Oracle healthcare flagship Master Data Management product OHMPI, J2EE application. Provided leadership in architecture and development of OHMPI product covering features such as probabilistic dynamic matching, bulk matching updates & loading, provider Index, performance benchmarks, and engaging with product management and consulting division in prioritization of new features.
  • Security point of contact for OHMPI product. Additionally in the role of Project manager, managed other projects for the team as well.

Software Architect/Team Leader

Ironhawk Technologies
07.2009 - 08.2010
  • Lead, Architect and develop Federated Data Synchronization framework to be used by US Army to connect battle field platforms with National Highway.

Lead Software Architect

SUN MICROSYSTEMS
07.2002 - 04.2009
  • Technical Leader for team of Engineers in architecture and development/SDLC of Master Data Management (MDM)/Identity Management services, tools and applications using open standards Java, JavaEE, JDBC, BPEL, SQL, XML, JSP, Ajax, Web Services, Spring, NetBeans, UML and Design patterns methodology. Performed technical reviews and mentoring. Provided architecture guidance to internal and external communities. Collaborated with product management in roadmap planning and requirement analysis. Educated partners, professional services specialists at various training events. Main project
  • Architected and co-developed high performance, scalable, distributed, cluster based multi-threaded bulk probabilistic matcher and data loader, ensuring company of competitive advantage in high performance matching and data loading. Patent.
  • Leader of team that architected and implemented multi domain manager that interacts with multiple Master indexes deployed as EJBs in multiple Application Servers and provides relationships and hierarchies, enabling Sun MDM suite in new markets.

Principal Member of Technical Staff

ORACLE CORPORATION
05.1995 - 05.2002
  • Designed and implemented various full stack projects.

Education

MS - Computer Science

University of Kentucky
Lexington, KY

BS - Electrical Engineering

Punjab Engineering College

Skills

  • Languages: Java, Scala, Python, C/C, SQL
  • Distributed Technology: Hadoop, Spark, AWS, S3, Kubernetes, Docker, Parquet, JavaEE, Microservices, JMS, REST
  • Domain: Risk Management, Ad Tech, Master Data Management, Health Care
  • Databases: Oracle, Postgres, MySQL, SQL Server, Redshift, MongoDB, Kudu

PATENTS

  • 1. Method and System for Distributed Bulk Matching and Loading, Oracle (Patent # 8943057, sole inventor)
  • 2. Configurable Dynamic Matching System, Oracle (Patent # 8849837, sole inventor)
  • 3. Bulk Matching and Update, Oracle (Patent #2929567, sole inventor)
  • 4. System and Methods for Dominant Attribute Analysis, Criteo (Patent pending)
  • 5. Semantic model based targeted search advertising, Criteo (Patent pending)

Timeline

Principal Engineer

Moody’s Corp.
08.2018 - Current

Architect

Criteo
03.2015 - 07.2018

Principal Engineer

Datapop Inc
04.2012 - 03.2015

Senior Principal Engineer/Lead Architect

Oracle Corporation
07.2009 - 04.2012

Software Architect/Team Leader

Ironhawk Technologies
07.2009 - 08.2010

Lead Software Architect

SUN MICROSYSTEMS
07.2002 - 04.2009

Principal Member of Technical Staff

ORACLE CORPORATION
05.1995 - 05.2002

BS - Electrical Engineering

Punjab Engineering College

MS - Computer Science

University of Kentucky

Similar Profiles

CREATE PROFILE