Pooja Balusani

Experience

• 

  Noodle.ai

  May 2017 - Jul 2017

  Data Science and Software Engineering Intern

  Built an Automated Time Series Application that provides in depth Exploratory Data Analysis of input data. Applying Statistical Tests such as Augmented Dickey-Fuller test for finding stationarity of a time series, generating Time Series Plot, STL Decomposition, Cross-Correlation and providing forecasts using Facebook’s Prophet

• 

  Noodle.ai

  Jan 2018 - May 2018

  Data Science and Product Development Intern

  Designed and Developed Reusable end-to-end AI pipeline in python and spark, highlighting a variety of automated visualization, feature engineering and modeling techniques such as Clustering, Classification, Regression and Survival Analysis coupled with various evaluation metrics

• 

  Noodle.ai

  Jul 2018 - Feb 2023

  Worked on Asset Flow Product where I, • Detected anomalies and failures for different assets using millisecond-based sensor data • Developed a reusable Artificial Intelligence pipeline using clustering and classification techniques in Python to detect and predict anomalies along with model interpretability techniques• Enhanced anomalies generated by building a drift detection and model monitoring module resulting in a 15% reduction in equipment downtime Worked on Product Quality Flow Application where I,• Developed a module to predict downstream product quality, identify drivers of quality deviations and recommend optimal controllable process settings to reduce those deviations to produce desired grade of steel• Implemented regression modelling techniques such as xgboost and random forest along with model interpretability mechanisms such as SHAP, LIME to highlight feature contribution resulting in a 20% decrease in quality variability Worked on a Demand Forecasting Application where I, • Developed a module providing an estimate of future demand on multivariate timeseries data along with external features• Comparing baseline models (naïve, ses), and deep learning (seq2seq) models resulting in 15% decrease in safety stock levels and 17% improvement in fill rate with more accurate and granular forecasts

  • Senior Data Scientist

    Apr 2021 - Feb 2023

  • Data Scientist

    Oct 2019 - Mar 2021

  • Associate Data Scientist

    Jul 2018 - Sept 2019
• 

  PwC Middle East

  Mar 2023 - now

  Senior Associate Technology Consultant - Data & AI

  • Led the development of a scalable AI Toolkit adopted by 16 countries, refining AI assessment & adoption strategies.• Built an AI-driven investment platform for a Saudi government entity, improving decision-making efficiency by 30%.• Developed a Gen AI MVP automating funding request review tool, achieving 75% accuracy and reducing processing time.