Deepika N Email & Phone Number

Chennai, Tamil Nadu, India

As a Data Scientist in my current role, Spearheaded the development of a search and chatbot system leveraging Elasticsearch and LLMs to streamline document retrieval and improve user interaction, enhancing operational efficiency by 30%. Automated the ingestion of large volumes of documents from AWS S3 using PDFPlumber, PyMuPDF, and AWS Textract, reducing manual handling by 80%. Developed complex algorithms to convert documents into embeddings and implement cosine similarity and KNN for accurate document ranking in Elasticsearch. Implemented a Retrieval-Augmented Generation (RAG) system and multi-threading to optimize search performance, reducing query response times. Integrated LLMs for generating precise responses to user queries, improving engagement by 25%. Enhanced the chatbot's memory to retain context, improving accuracy in multi-step user queries. Built real-time communication capabilities through socket programming, delivering a responsive user experience. Processed form data using computer vision techniques like SIFT and fuzzy matching to convert hardcopy forms into digital formats and generate actionable insights. Led the creation of dynamic data visualizations to empower users with in-depth analysis and comparative insights.

Completed 6 out of 8 projects which involves : Project1: Housing in Mexico ✓Load data into a pandas DataFrame ✓ Clean and combine messy data ✓Examine the relationship between variables using correlation. ✓Create insightful visualizations like histograms, scatter plots, and bar charts.Project 2:Housing in Buenos Aires ✓Load multiple CSV files into a pandas DataFrame ✓Clean messy data with a wrangle function ✓Create a machine learning pipeline with feature encoding and imputation ✓Create insightful visualizations like a Mapbox scatter plot, a heatmap, and a bar chart for model coefficientsProject 3: Air Quality in Nairobi ✓Get data from a MongoDB database ✓Explore and clean time series data. ✓Build autoregression models. ✓Tune hyperparameters.Project 4: Earthquake Damage in Nepal ✓Extract data from a SQL database ✓Perform a randomized train-test split ✓Build a logistic regression model for classification ✓Build a decision tree model for classification ✓Tune model hyperparametersProject 5: Bankruptcy in Poland ✓Navigate a file system using the command line ✓Load and save files using Python ✓Work with imbalanced data using resampling ✓Build ensemble models like random forest and gradient boosting trees ✓Evaluate a classification model using precision and recallProject 6: Customer Segmentation in US ✓Conduct complex exploratory data analysis on one of the biggest datasets in the DS lab, and compare sub-populations using a side-by-side bar chart. ✓Apply key concepts of unsupervised learning into practice by building an k-means model. ✓Conduct feature selection for clustering based on variance. ✓Reduce high-dimensional data using principal component analysis (PCA). ✓Design, build and deploy a Dash web application.

India

Chennai, Tamil Nadu, India

Masters, Mathematics

Ug, Mathematics