About Swayonok kundu
-
Viewed 869
About me
Actively seeking out for full time roles in Data Science domain. Have around 6+ years of industry experience in Analytics, predictive modelling, and visualizations, manifesting insightful stories from business problems. Delivered end-to-end projects which necessitates statistical modeling algorithms, predictive modelling, supervised and unsupervised machine learning models in Python and R across domains of finance, retail, healthcare, marketing and supply chain.
Data Analysis and Statistical Model Building: Python (Pandas, Scikit-Learn, matplotlib), R, SQL, Alteryx, MS Office
Methodologies: Regression, Classification, Clustering, Machine learning, Neural Network, Forecasting, NLP, PCA
Business Intelligence Tools and Deployment: Tableau, Power BI, AWS
Data and Product Management: MSSQL, SQLite, SQL Server, Jira, Confluence
Education
-
- Present
University at Buffalo
Masters in Data science
-
- 2013
University of Kalyani
Master of Science, Statistics
-
- 2010
University of Calcutta
Bachelor of Science, Statistics
Experience
-
2022 - 2023
Lenovo
Data Scientist
• Spearheaded a project to predict the repeat repair rate, cost per claim, parts used for service orders of machines leveraging statistical techniques and Machine learning models for all Lenovo brands across globe.
• Transmuted the findings for all KPIs through Power BI dashboard.
• Worked in Topic modelling for unstructured data with customer service requests.
• Worked in Geo-spatial analysis with customer install base.
-
2019 - 2020
Fidelity Investments
Lead
• Managed a project of forecasting for demand hours and volumes of resources attending clients for Defined benefit plan which helped the managers to allocate their respective team resource accordingly. Resource planning
helped to brought down the hiring cost by 50%. -
2016 - 2019
Accenture
Consultant
• Led a project in workforce labor forecasting and Optimization to forecast hotel traffic for a hospitality management company using traditional Time- Series and Machine-learning models, predicting human traffic.
• Gathered internal and external predictor variables, performed all the EDA, scraped weather and events data.
• Built multivariate Time-Series models with Arima-x, UCM and Ensemble methods of all bagging, boosting models
to achieve an accuracy of 98% at highest level and granular level accuracy of 94% on average.• Worked with a retail giant of South-East Asia in a demand planning project to optimize the inventory and stock outs by sales prediction.
• Built a credit risk model to predict probability of loan applicants who will default on their loans using Logistic
regression to facilitate client to make decisive actions based on the model, lowering the bad loans by 20%.• Transmuted actionable insights to stakeholders by creating dashboards in Tableau.
-
2014 - 2016
Nielsen
Data Analyst
• Worked in a demand forecasting project for volume of hypermarkets for fast moving consumer goods at diverse levels of SKU and automate processes in R for all SKUs and optimizing the inventory management to reduce the
wastage cost by 10% on average per SKU.