Prattay Chakrabarty

About Prattay Chakrabarty

As a Data Scientist with 5 years of experience, I have honed my expertise in Python, R, and SQL, specializing in Regression, Market Mix Modeling, Price & Promotion,Time Series analysis, Demand Forecasting,Churn/Attrition Modeling,Generative AI & Agentic AI. My career includes impactful engagements with prominent Consumer Markets, Healthcare, Technology, Banking & Insurance companies, where I have consistently delivered data-driven insights to optimize marketing strategies, Retention, Revenue Growth and various business operations.Some of the problem Statement worked on :• ML-based Price & Promo for F&B to boost revenue. • AI-driven client prioritization for a Fortune 6 healthcare, enabling strategic moves. • Spend forecasting for a chipset giant, optimizing budgets. • Mixed-Effect for a beverage brand, maximizing ROI. • Bayesian for FMCG, leveraging priors for media impact. • Vendor risk prediction for a CPG to predict & prevent supply disruptions.• Fraud Detection Solution for a Leading US based Beauty Brand to reduce False Positive • Product Information Enrichment for a US based retailer using ML & Gen AI • ML Based Consumer Attrition Modeling & Next Best Action using Gen AI for a US Based Bank• Document Intelligence with Generative AI & vector DBs for text analytics & Chat Bot. • HR Intelligence with auto-JDs & resume analysis, guarded by custom measures. • DataGenAI pipeline (LLMs + GANs) for synthetic data, preserving privacy. • AI-based data classification & hidden relationship detection. • Multiple Domain-agnostic ML accelerators for rapid modeling & forecasting. Algorithm / Tools applied in these projects : Machine Learning : Regression, Classification, Clustering & Anomaly Detection ( Linear Regression - Additive & Multiplicative, Lasso-Ridge-ElasticNet Regularization, Mixed Effect Modeling,Logistic Regression, Decision Tree & Random Forest, XGBoost, CatBoost, LightGBM, K-Means - General & Constrained, K-Mode, K-Prototype, Isolation Forest, OneClassSVM, LOF )Time Series : ARIMA, SARIMA, SARIMAX, Prophet, ETS, Croston, IMPAP, MFLES, TFT & Custom Ensembled Models.Gen AI : LLM\'s ( OpenAI & Open Sourced ), Agentic AI, Hugging Face, CrewAI, LangChain,Pinecone, FAISS, Azure AI, Guardrails.Others : Python, R, SQL, MS-Excel, Azure, Databricks, Snowflake, Streamlit.

Experience

1. Data Scientist

   PwC Acceleration Center India

   Jul 2024 — Present · Kolkata, IN

   Delivered Price & Promo Solution for a Large F&B across multiple markets • Built AI / ML Based Prioritisation Model for a Fortune 6 Healthcare Client• Developed ML Based Spend Forecast Framework for one of the Largest Chipset Manufacturer • Fraud Detection Solution for a Leading US based Beauty Brand to reduce False Positive • Revenue forecasting for 1400+ Retail Store to Optimize StoreOps Planning• ML & Gen AI Based Product Attribution Solution Deployed to drive 100 Million Revenue for E-Com• Agentic Framework Developed to transform Marketing Value chain of CPG & Retail• Created Multiple domain agnostic AI/ML accelerator for Modelling, Forecasting & Agentic AI Workflows.• Contributed on Multiple RFP & PoV on AI,ML,GenAI & Agentic AI across Domains

Education

• Ramakrishna Mission Boy's Home High School

  High School, Computer Science

  2015 — 2017

• Techno India Saltlake

  Bachelor of Technology - BTech, Information Technology

  2017 — 2021