Mar 2024 – Jun 2025 · 1 yr 3 mos · Bengaluru, Karnataka, India · Hybrid

• Demand Forecasting:
• Developed a forecasting solution for a leading apparel and footwear client using statistical models (MA, ARIMA, SARIMA, STLF, Prophet, Exponential Smoothing), achieving 71% accuracy.
• Delivered a bottom-up forecasting solution for 250k unique planning-level intersections using a hybrid of State Space Models, Hierarchical Clustering, and LightGBM, significantly improving forecast accuracy and reducing bias.
• Ensured end-to-end project success through unit testing, regression testing, and coordinated UAT/SIT, culminating in smooth production deployment and stable batch runs.
• Size Profile Optimization:
• Built a Size Profile Optimizer with dynamic sales correction strategies to improve allocation and sizing accuracy for a leading fashion retailer.
• Designed an anomaly detection framework leveraging ranking and segmentation logic to flag size profile inconsistencies and guide demand adjustments.
• Other Achievements:
• Recipient of the Execution Award for high-quality delivery on critical data science projects.
• Top 3 in o9 organized EDA Hackathon for delivering deep analytical insights.
• Earned o9 Knowledge Master – Demand Planning certification.
• Technical skills:
• ARIMA, SARIMA, STLF, Prophet, Exponential Smoothing, Xgboost, LightGBM, Ranking, Segmentation, Size Profiling, Anomaly Detection, Statsmodels, Nixtla, Unit & Regression Testing, Data Visualization (Matplotlib, Seaborn), Python, Pyspark, DuckDB, SQL, IBPL

Jul 2021 – Mar 2024 · 2 yrs 8 mos · Bengaluru, Karnataka, India · Hybrid

• Demand Forecasting & Optimization:
• Built in-season forecasting solution using ARIMA, SARIMA, SARIMAX (time-series models), ERT, RF, ES, MA, LSTM, and GRU.
• Developed a next-generation inventory planning solution enabled by Mixed-Integer Linear Programming and greedy heuristic-based models
• Built the API endpoints using FastAPI and hosted it on Azure App Service
• Increased the client revenue by 1% and reduced the stock-outs by 40%
• Worked on process automation by writing Python scripts to crawl the actual and prediction data from S3 and deployed the module in the existing Airflow
• Worked on Athena and Redshift for data validation and debugging
• NLP:
• Completed an NLP-based project where the task involved checking the missing contractions, headings not starting with actionable words and presence of past or future tense in a sentence.
• Built a UI using Streamlit, to give the status of checks for given inputs
• Built an interactive PDF chatbot using ConversationalRetrievalChain from langchain.
• Summarized the pdf with load summarize chain from langchain.
• Built a streamlit UI to upload PDFs, choose models, and to interact with Chatbot
• CV:
• Completed an end-to-end computer vision project to distinguish between the real image and diagrams of products on a website and built a UI using the Streamlit framework.
• Built the stock monitoring system to have a continuous update of inventory levels of shelf products using advanced computer vision models which will help in on-time stock replenishment
• Applied transfer learning by training a YOLOv5 model on a custom dataset which was annotated using Roboflow and used Gradio library for UI
• Skills:
• Deep learning •ML • Gen AI • LLMs • Supply Chain • Supply Chain • Inventory optimization • AWS • Azure • Databricks • Fast API • Streamlit • Python • Pyspark • SQL • R • Jira • Snowflake • Confluence • MSTR • PYOMO • ORTOOLS
• Awards:
• Start Award (Wizard @ Work)
• 3x Client Value Creation Award
• Stewardship Award

Jun 2020 – Aug 2020 · 2 mos · Bengaluru, Karnataka, India · On-site

• CSIR Summer Research Programme 2020:
• Conducted multivariate linear regression (MLR) to model the relationship between aerosol optical depth (AOD) and precipitation.
• Processed satellite datasets (MODIS) and ground-based meteorological data for climate variable extraction.
• Utilized Python (NumPy, Pandas, Scikit-learn, Matplotlib) for data analysis, statistical modeling, and visualization.
• Applied GIS tools (ArcGIS/QGIS) for geospatial mapping of aerosol distribution and rainfall patterns.
• Performed statistical analysis (correlation analysis, PCA, hypothesis testing) to interpret climate-pollution linkages.
• Investigated the impact of air pollution on rainfall trends over Kanpur, contributing to climate change studies.
• Key terms: ClimateChange, AirPollution, GIS, Rainfall, MultivariateRegressionAnalysis, StatisticalAnalysis