Jun 2021 – Mar 2024 · 2 yrs 9 mos · Pune, Maharashtra, India · On-site

• Built sophisticated AI-driven data pipeline, streamlining the audit process for expenses, invoices, and card transactions. This system offers financial teams a comprehensive, 360-degree perspective on their spending patterns. Beyond providing detailed visibility into expenditures, it also benchmarks our performance metrics against those of our industry counterparts. This innovative approach ensures a thorough and competitive financial analysis, enhancing our capacity to make strategic, data-driven decisions.
• AppZen is the first artificial intelligence (AI) solution for back office automation. AppZen’s platform uses AI to automate expense report auditing and instantly detect compliance issues and fraud. AppZen’s patented AI combines computer vision, deep learning, and natural language processing to automatically read and understand expense reports, receipts, and travel documents and cross-check them with hundreds of data sources in real time to determine the accuracy and legitimacy of every expense. This enables companies to detect fraud and compliance issues in seconds.

Jun 2019 – May 2021 · 1 yr 11 mos · Hybrid

• Worked with clients such as Moet Hennessy (NY,USA) [subsidiary of Louis Vuitton], Koch Industries (Atlanta , USA) , Kearney (Chicago , Illinois) and cello capital (NY, USA) to help them in building data platform.
• Cello Analytics:
• Worked on a project involving CDU processing, which fed data into Redshift and employed a Spark process on an EMR cluster. Output was generated in both JSON and Parquet formats, which was then consumed by DynamoDB and Redshift for further processing. The entire EMR process was automated, enabling the launch of multiple clusters with an auto-termination feature to minimize costs.
• Koch:
• The primary objective was to build a data pipeline for IoT data, with an incoming data volume of about 2TB daily. The project involved converting old architecture to a new one to process data faster with 99.99% accuracy using PySpark and EMR clusters. Employed PySpark DataFrame architecture to convert data from unformatted JSON files to Parquet file format. Managed to process 100GB of data within 10 minutes using 20-node EMR clusters.
• LVMH – NY:
• Managed multiple data migration verticals with a team in India, specifically for data feeds from Salesforce and Analplan to Oracle and SQL Server. Prepared architecture for making AWS instances a big data platform, accommodating multiple data feeds such as sales data, weather data, social data, Salesforce data.
• Frena Analytics
• Started data analytics firm and helped startup teams in getting up and running
• Explored various startups opportunities in India in multiple segments and identified risks associated with
• it. Mentored budding entrepreneurs and Helped startups companies by analyzing market data to make strategic decisions about their strategy on expansion, funding allocation, building product, setting up new shops or venturing into new market , this helped companies to achieve their goals and able to expand well even in hostile environment
• Startups I worked with are buzzingalab , washbucket,,WeTrade

Jun 2018 – Jul 2019 · 1 yr 1 mo · Gurgaon, India

• I have collaborated with teams to extract, store, and analyze large datasets from our proprietary GSIM/SA analytics platform, utilizing advanced analytical queries such as session, partition by, cube, and rollup. I am proficient in extracting raw data from Hadoop Analytics via Hive Queries.
• My work has included an in-depth market analysis within the Single-Lens Reflex (SLR) camera segment, examining various features, settings, and modes over a two-month period. This was accomplished using both GSIM/SA and Hadoop raw data.
• My role extends to business strategy and planning, with a particular emphasis on implementation and execution. I have spearheaded multiple strategy and transformation projects, predominantly within the camera segment.
• I have developed automation methods to monitor new features specific to the Indian market, such as Chat-over-video, Social camera, and Dual Camera.
• In addition, I have leveraged advanced analytics to understand 3rd party app usage across various regions and operators in India, extracting GSIM data from Hadoop analytics using Hive queries.
• I adopt both top-down and bottom-up approaches in meeting business goals with data analytics, ensuring a comprehensive strategy for our data-driven initiatives.

Feb 2017 – May 2018 · 1 yr 3 mos · San JoseI

• Participated in a significant data migration project at Equinix, a global leader in colocation data center services.
• Led the design and architecture of comprehensive solutions for migrating all backend databases from Oracle and MongoDB to Postgres, based on a deep understanding of business processes and release schedules.
• Developed an end-to-end solution leveraging a Python and SQL programming framework.
• Engineered an ETL framework using Python, with SQL responsible for all data processing activities.
• Utilized Spark as the primary tool for data processing in a clustered environment.

Apr 2015 – Nov 2016 · 1 yr 7 mos · San Mateo, California

• Customer Collaboration: Collaborated extensively with both internal and external customers to optimize the understanding and utilization of our comprehensive cloud platform, which includes AWS services, Hadoop as our Big Data platform, Vertica, and various reporting and data warehousing tools (Tableau, Big Data Discovery, D3, Oracle & Postgres), along with ETL (Talend, Python).
• Mongo Hadoop-Vertica Integration: Led the initiative to integrate data from Glip, a social media company acquired by RNG. Successfully managed the migration of data hosted on Mongo servers to a Hadoop cluster and subsequently pushed it to a Vertica cluster for reporting purposes. This comprehensive project involved architecture, design, and development. Leveraged JSON UDF to extract data from the JSON format to the Hadoop cluster and utilized Spark stream and Spark SQL streams for data pipeline creation.
• Platform Reporting (Hadoop-Oracle-Web-Vertica): Spearheaded the integration of all platform reporting data from Hadoop with customer data from Oracle and live API data from the web. Consolidated data was pushed to the Vertica platform on a daily basis for reporting.
• ETL Migration: Successfully transitioned all Talend jobs to a homegrown Python framework, improving efficiency and customizability in our ETL processes.

May 2013 – Apr 2015 · 1 yr 11 mos · San Francisco Bay Area

• Next Gen Data Science Warehouse/BI Architecture: Led a team of 5-7 professionals in building a state-of-the-art Data Science warehouse and BI architecture.
• ETL Framework and Script Revamp: Orchestrated the design of a comprehensive ETL framework, centralizing all ETL scripts into a unified repository. This facilitated centralized logging, monitoring, and simplified maintenance and reporting. Led a thorough analysis and overhaul of individual ETL scripts in collaboration with the team, resulting in a significant reduction in ETL load time from 15 hours to just 3 hours.
• Qlikview Administration: Oversaw the administration of Qlikview, an OLAP reporting analytics tool, and guided developers in creating more efficient reports through query optimization and improved data architecture.
• Amazon Web Services: Well-versed in Amazon Web Services, including Amazon EC2, Amazon S3, Amazon SimpleDB, Amazon RDS, Amazon Elastic Load Balancing, Amazon SQS, and other AWS services.

Mar 2009 – May 2013 · 4 yrs 2 mos · Palo Alto , CA

• Vmware Acquired Packetmotion in 2011.
• Responsible for all data integration aspects of this product, including data analysis, migration, replication, and
• business intelligence
• Created design document for data integration and migration for various data sources to build an integrated
• application environment
• Worked with Product Managers and Solution Architects to understand the business objectives for data
• integration projects and architected the OLTP and OLAP model
• Worked with business analytics to come up with all end-user requirements; modified data model and
• optimized query design accordingly
• Working on porting code from Oracle to PostgreSQL and MySQL
• Using Mongodb as front end for Query Caching
• Evaluating Hadoop as Big Data Solution for the replacement of traditional RDBMS in combination with
• Postgres. Used various Hadoop features, such as Map/reduce, Hbase, Hive, Pig, etc. Installed and configured
• 10-node Hadoop cluster and used Hive queries to access data.
• Used VMware’s in-house Serengeti tool for deploying and managing various distro
• As a part of a team, migrated all existing PacketSentry infrastructure to a virtualized environment
• Installation and configuration experience with ESX server
• Hands-on experience with various VMware products, including Vcenter, Vsphere, Vconvertor, and Workstation
• Worked as Lab Proctor for infrastructure labs during hands-on labs at VMworld 2012. At VMworld, we ran
• 9,080 labs and deployed over 110,000 VMs.
• My team is solely responsible for all information, database, and BI-related activities, including data modeling, ETL,
• OLTP, data warehouse, and reporting.
• o I am primarily involved with managing the team as well as hands-on experience with every critical aspect of
• projects
• Notable projects include:
• Streamlining the data archiving process
• With an embedded database, we need to make sure that data are purged on their own, based on various
• requirements (e.g., scheduled, space, corrupt data, ad-hoc).