Anuj Khot

Designed and implemented a Chatbot using Langchain framework powered by LLM and OPENAI GPT 3.5 Model to enhance proactive customer interaction and lead generation for the marketing team. Utilized FastAPI for efficient response streaming, Deployed it on Azure cloud using Azure Web App and storing conversation history in Azure DB. Developed and Deployed a Q&A system over structured data for a finance client using SQLDatabaseChain and SQL Agents to retrieve results from the database, presenting them in two formats: a. Data interpretation insights b. Visuals(Charts). Developed and executed customer segmentation strategy using RFM analysis and BG/NBD model, resulting in identification of 5 distinct customer segments with varying purchase behaviors. Built Power BI dashboards to visualize core business KPIs (weekly revenue, sales, profit) and generate insights which increased efficiency by 50%. Created Time Series Forecasting models such as ARIMA and XGBoost for forecasting sales. Scrubbed, processed and analyzed large-scale datasets of upto 2000 Mb from multiple sources, leveraging pandas and numpy techniques to identify data trends, discover patterns and draw conclusions.

Worked on developing a CNN model to detect and compare ideal stances to stances being played by a badminton player, resulting in improved accuracy of 90%, reduced false positives by 40% and increased the speed of detection by 60%.