Sweta Mangaraj

Aspiring data scientist with a robust background in machine learning, statistical analysis, and predictive modeling. Skilled in translating complex datasets into actionable insights that drive decision-making and business strategy improvements. Demonstrates strong problem-solving abilities and mastery of Python, R, SQL, and data visualization tools. Previous work has led to significant enhancements in operational efficiency and revenue growth through data-driven strategies.

• Consumer Complaints Resolution: Consumer complaint resolution is important to any
  business. We had details of over 6,00,000 consumer complaints from a BPO which worked
  with various global banks along with whether the consumer disputed the resolution provided
  by the bank or not. The client, the BPO wanted to gain better understanding of why certain
  resolutions were disputed and also wanted to be able to predict whether the consumer will
  dispute a resolution or not. This was so that more attention can be given to how the
  complaints are handled as well as how to convincingly convey the final resolution tothem.
  Solution- We parsed the text data using Python and then created a machine learning
  model using Naïve Bayes to predict resolution disputes.
  Over the next 6 months, disputes were brought down from 34% to 19% of all complaints.
• Pharma: A mid-sized pharma company was facing issues with counterfeits of its brands.
  It wanted to target and put these counterfeit operators out of business with the help of the
  government. In order to make maximum impact it wanted to focus on high value operators
  for the time being. To do that it needed to predict sales of counterfeit medicines and
  counterfeit operators given the data collected from various sources about their operations
  and sales so far so as to understand their operations sizegoing forward.
  Solution- Created a Random Forest model to predict counterfeit medicine sales and
  pinpoint high value operators. Communicated the findings to the client for further action.
  
• Insurance cross-selling: A general insurance client wanted to know whether an existing
  policy holder would be interested in a new car insurance policy. Information about the
  existing policy holders consisted of 86 variables and included product usage data and
  socio-demographic data derived from zip area codes.
  Solution- Used XGBoost to create a ML model to predict those customers most likely
  to respond positively to the new product and the marketing campaign.
  Model had an accuracy of 80% and increased campaign ROI by 68%.

• Certified machine learning with Python
• Certified data analyst using SQL
• Certified business analytics with R
• Certified in deep learning

• Actively volunteer with the Art of Living Foundation.
• Completed courses in meditation, breathwork, and yoga.
• Involved in multiple seva projects for societal improvement.
• Organized courses focusing on mental and physical health.
• Contributed to blood donation camps and health checkups.
• Engaged as an active member in animal welfare through Ekamra.