Prakshi Karkera | Portfolio

Work Experience

Data Analyst Intern

Dun & Bradstreet India

July 2025 - Present

Developing Total Addressable Market (TAM) models to optimize commercial targeting strategies.
Writing and deploying scalable PySpark code within Databricks environments for large-scale data manipulation.
Collaborating closely on high-impact client deliverables and automated business intelligence pipelines.
Identified Total Addressable Market (TAM) and customer segments using RFM analysis, improving targeting accuracy for commercial strategy.
Reused existing workflows and templates to cut manual effort and turnaround time on TAM analysis by 30%.
Built and optimized 50+ interactive dashboards using PySpark, SQL, and Excel, supporting business insights across sales, risk, and ESG.
Partnered with cross-functional teams to translate business requirements into actionable reporting solutions.
Supported client delivery and presales with data-driven insights and business presentations.
Automated data extraction and reporting pipelines using Power Automate and Microsoft Copilot, reducing manual intervention.

Skills & Capabilities

From raw data to deployed insight — the tools and stacks I build with.

📊

Data Analysis & Visualization

Turning raw, messy datasets into clean, decision-ready insight through exploratory analysis and dashboarding.

PythonPandasPower BITableauDAX

🤖

Machine Learning & Gen AI

Building classifiers, RAG pipelines, and LLM-powered tools that respond intelligently to real data.

Machine LearningDeep LearningTransformersFAISSGroq API

🗄️

Databases & Big Data

Writing and deploying scalable data pipelines for large-scale manipulation and storage.

MySQLPySparkDatabricksOpenPyXL

🧩

Application Development

Shipping full desktop and web interfaces around the data — not just the analysis behind it.

StreamlitCustomTkinterLangchainSMTP

📈

BI & Reporting

Designing interactive dashboards that turn raw datasets into actionable visual insight.

Power BITableauDAX ModelingReporting

🧠

NLP & Text Intelligence

Applying language models and semantic search to classify, retrieve, and generate from text.

Hugging FaceBERT ClassifierSentence TransformersRAG

Featured Projects

★ RESOLVEAI

AI Complaint Classifier and Responder

RESOLVEAI is an AI-powered customer complaint management system built using Streamlit. It intelligently classifies complaints by category and urgency, generates AI-based responses, and provides separate interfaces for customers and administrators to manage and resolve issues effectively.

PythonMySQLStreamlitHugging FaceTransformers Groq APIBERT classifierFAISS

★ Bank Application

Integrated Banking Solution

A desktop-based Banking Application built with Python and CustomTkinter, featuring account management, transactions, and customer verification connected to Power BI for real-time analytics and reporting, enabling data-driven decision-making for banking operations.

PythonMySQLTkinterCustomTkinter Power BIPandasSMTPOpenPyXLRandom

★ PDF QnA

AI-powered PDF question-answering system

PDF_QnA_RAG is an intelligent question-answering system that allows users to upload a PDF file and interact with its content using Retrieval-Augmented Generation (RAG). It combines document chunking, semantic search, and large language model responses to generate accurate, context-aware answers.

PythonPyPDFLoaderSentenceTransformers Groq APILangchainFAISSGemma-2b-it model

★ Webscraping

Flipkart Product Data Extraction and Preprocessing

Flipkart Product Data Extraction and Preprocessing is a project that focuses on scraping product information from Flipkart and performing necessary preprocessing steps to make the data suitable for analysis.

PythonBeautifulSoupPandas NumPyScikit-learn

★ Prediction

Placement Prediction App

Live Demo

Placement Prediction App is a machine learning project that predicts student placement outcomes based on their academic performance and other features.