EXL Analytics | Senior Business Analyst

EXL Analytics (erstwhile Inductis) is a global professional services firm that helps large companies leverage the information age to make data-driven decisions through deep analytics.
place Gurgaon date_range August 2015 to Present
  • Adaptive Content Extraction from legal documents:
  • The objective of the problem is the reduction of manual work for a US Based legal consulting client in processing the huge amount of scanned unstructured legal documents.
  • Content Extraction from scanned logistics and transportation bills:
  • Developing a product prototype to extract shipping and billing information from scanned images of bills for a leading logistics company in the US.
    Key Technologies being used: Python, OpenCV, Tesseract-OCR.
  • Real-time analytics for IoT devices:
  • Created the automated architecture for real-time data processing for the IoT devices on Amazon web services using Python. Also, developed & rendered a responsive web application using Flask (Python) which provided analytics supported insights and warnings to the customers according to their usage data patterns. The prototype was implemented with smart meter data for a utilities-based client.
    Key Technologies Used: Python, D3js, Flask, PostgreSQL, Sqoop, Boto3 (AWS API).
  • Image Classification to identify distracted drivers:
  • Mentored a team of interns to work towards developing an image classification framework. The objective of the problem was to classify car driver images into 10 categories ranging from safe driving to distracted driver (talking on phone, holding coffee cup, et cetera). We implemented CNN to train the images to respective classes and the result was validated with the data provided using log loss score. The solution obtained could be leveraged in the insurance industry to segregate the false claims from the genuine ones.
  • Social Media Influencer module:
  • Mentored an intern in developing a social media influencer tool to identify and assess the impact of an influencer present in a social network. A twitter network of multiple news sources was observed and a Centrality and Superfan Score was assigned through topically sensitive tweet tree of users. The solution can be used to maximise marketing-content reach with minimum resources.
  • Misuse Detection for Online Subscription based client:
  • Modeled a fraud detection architecture to identify the non-compliant users subscribing to journals of a healthcare client using concurrency analysis, usage pattern analysis and statistical outlier analysis methods. The architecture was end-to-end automated and was hosted on AWS with PostgreSQL as the data warehouse. This architecture had the capability to provide real-time analytics to the client and could prevent the misuse of their service or product by flagging off the non-compliant users. The modules were developed using Apache Spark framework and were implemented in Python.
    Key Technologies Used: PySpark, PostgreSQL, Sqoop, Boto3 (AWS API).
  • Python API for Retrieval and Analysis of Data:
  • Developed a re-usable Python API for retrieval and analysis of data from multiple sources for a health-care client based in the US. This enabled the team to render quick and easy visuals on the front-end of the prototype.
  • Travel & Expense Compliance Tool:
  • Developed a complete responsive web-based tool using Python and D3js charting library for a leading US Credit card company to monitor the travel & expense compliance.
    Key Technologies Used: Flask, Python, D3js.

The Custom-T | Co-Founder

A merchandising solutions and design company.
place Delhi date_range March 2012 to May 2015
  • Responsible for web development, inbound-marketing, social-media marketing and data management.
  • Analyzed sales records, derived trends and identified high-growth sectors for the sales & marketing team.
  • Dealt with over 200 clients and shipped 50000+ pieces in a span of 3 years. Clientele include Adobe, Tata Communications, Genpact, Ibibo Group, Findable, IIT Delhi, St. Stephens College, FMS and many more.

Khaana On The Go | Intern

Khaana On The Go was a Food Delivery cum Restaurant Search based web startup in Gurgaon.
place Delhi date_range June 2014 to July 2014
  • Analyzed potential keywords and meta-tags to improve the search visibility and indexing of the website.
  • Was a member of the development team of www.khaanaonthego.in and achieved UX rating of 99% on Google.
  • Developed web application for the location services integration using Google Maps API and HTML5 Geolocation.

Indian Oil Corporation Limited | Summer Trainee

IOCL is an Indian state-owned oil and gas corporation.
place Delhi date_range June 2013 to July 2013
  • Responsible for optimizing crucial parameters in the feasibility report of a Poly-propylene plant in Orissa.
  • Analyzed various routes of project planning and implementation of the Poly-propylene plant and refined the report.

Indian Army 509 Base Workshop | Trainee

place Agra date_range December 2012 to January 2013
  • Learnt the different processes from receiving a work order to the mass production. Worked under the Methods Engineering and the Manufacture and Repair Departments.