> SYSTEM INITIALIZED_

Pranav Rebala_

[Data Analyst | ML Engineer]

Transforming data into actionable insights through Machine Learning, Analytics, and Full-Stack Development

SCROLL

// INTRODUCTION

AboutMe

With a strong foundation in computer science from PES University, I've honed my expertise in Python, full-stack development, and data-driven decision-making. Having recently completed my Master's in Business Analytics at Babson College, I bring a unique blend of technical expertise and business acumen.

My hands-on experience includes developing scalable applications, designing innovative analytics solutions, and collaborating on projects that leverage data to drive impactful outcomes. Passionate about creating meaningful change, I aim to contribute to projects and teams by harnessing the power of technology and analytics to solve real-world problems and unlock new opportunities.

Name

Pranav Rao Rebala

Email

navrebala@gmail.com pranav.rebala11@gmail.com

Phone

+1 (617) 987-1268

Nationality

American

Languages

English, Hindi, Telugu, Kannada

Babson College

MS in Business Analytics

Completed

PES University

BS in Computer Science

Graduated

// TECHNICAL PROFICIENCY

TechStack

A comprehensive toolkit for building scalable applications and deriving insights from data

[Languages]

Python

JavaScript

C/C++

SQL

PHP

[Frameworks & Libraries]

React

Node.js

Tailwind

Pandas

Numpy

[Cloud & Tools]

AWS

Azure

Docker

MongoDB

Tableau

Jupyter

> Also experienced with: HTML5, CSS3, Matlab, Deluge, and more...

// ACADEMIC BACKGROUND

Education

Babson College

Master of Science in Business Analytics

2024 - 2025Wellesley, MA

PES University

Bachelor of Science in Computer Science

2020 - 2024Bangalore, India

// PROFESSIONAL JOURNEY

WorkExperience

Software Developer

Forfend Cybernatics

EngineeringSep 2022 - Jul 2025

▹Designed and deployed a LightFM-based hybrid recommendation model within a FastAPI microservice (Dockerized on AWS with RDS backend), predicting vehicle preferences from historical sales and behavioral data; handled 20,000 monthly kiosk and sales requests with 60% higher engagement and 22% higher conversion rates
▹Implemented a Generative AI-powered conversational chatbot with intent classification and LangChain-based RAG retrievers for appointment booking, inventory queries, and expert assistance in automobile dealerships; reducing customer support operational expenses by 50%
▹Developed an Agentic AI system for shop-floor management, integrating time-series forecasting (Prophet/XGBoost) and RL-based bay allocation to monitor live queue congestion across multiple service centers. Orchestrated decisions via a LangGraph agent and delivered insights through a live analytics dashboard with predictive workload optimization, improving operational efficiency by 60%

Full Stack Intern

Rezolve AI

InternshipJan 24 - Feb 24

▹Led a team to design and deploy real-time dashboards accessing Azure-hosted financial data for Wipro's BFS delivery vertical, enabling leadership to monitor key metrics for a 230,000+ employee global workforce
▹Implemented role-based access controls to ensure secure, compliant data visibility across user groups
▹Streamlined data ingestion workflows, reducing manual reporting time 40% and accelerating decision-making speed 30%

Full Stack Intern

Wipro Ltd

InternshipJul 2023 - Oct 2023

▹Collaborated and developed on an internal portal, utilizing React, Vite, Node.js/Express.js, SQL, and Azure
▹Significantly enhanced internal resource tracking, project utilization, and data-driven decision-making, boosting operational efficiency
▹Utilized React to develop a dynamic dashboard frontend. Used Axios for asynchronous data fetching. Used React Navigation for seamless user navigation. Leveraged Bootstrap to create an effective User Interface
▹Managed the deployment on Azure, ensuring high availability and scalability while integrating Azure Functions for dynamic scaling. The backend utilized MS SQL as the relational database, managed through SQL Server Management Studio
▹Integrated Role-Based Access Control through Azure Active Directory, ensuring Granular Access Control and secure data access

Data Science Intern

Flutura Business Solutions

InternshipJun 2022 - Aug 2022

▹Specialized in Engineering Workbench (EWB) Data Analytics at Flutura Business Solutions, conducting drone flight data analysis using EWB and Cerebra Solutions, extracting meaningful insights
▹Analyzed sensor data for drone failure cases using clustering, enhancing expertise in data analytics and digital twin concepts
▹Researched AR/VR tools, selected Vuforia, and integrated it with Unity to create a Drone AR environment. Simulated swarm scenarios with radar-zone detection and on-click individual drone analysis, showcasing practical skills

// PORTFOLIO

FeaturedProjects

Innovative solutions combining data science, machine learning, and full-stack development

// LATEST PROJECT

Late Edition

Football Analytics AIFootballBase

A real-time, multilingual RAG + LLM analytics assistant that converts raw sports datasets into conversational insights, charts, and exportable assets.

Why I built it

I wanted a single workflow that fuses data engineering, machine learning, and visual analytics so an analyst can ask a plain-language question and receive a data-backed answer, a chart, and an exportable slide in seconds.

How it works

Frontend: Next.js + React with a streaming UI that renders Chart.js visuals while receiving WebSocket updates from the backend.
Backend: FastAPI streams JSON and matplotlib charts; LangChain orchestrates tool calls, and OpenAI models perform intent classification plus slot filling.
Retrieval & Intelligence: FAISS with Sentence Transformer embeddings plus fuzzy matching for entity resolution; MCP coordinates tool invocations and data retrieval.
Data Ops: Pandas wrangles source data while a lightweight SQLite cache accelerates repeat questions. Everything is containerized with Docker and secured with Caddy for TLS/WebSocket passthrough.

What it delivers

Delivers concise, cited answers with inline sources, on-demand charts, PNG export, and automatic language detection across the top ten languages.

Real-world value

Research & finance teams get instant cited charts for investor decks.
Support teams generate knowledge-base assisted replies with confidence scores.
Retail/CPG teams receive promo lift estimates and SKU recommendations.
IoT/manufacturing teams gain anomaly explanations with prioritized inspection checklists.
Regulated industries turn policy documents into actionable product tasks and compliance steps.

VISIT SITE↗WATCH DEMO▶

If you work where ML, data engineering, and product intersect—or have thoughts about vertical use cases—I’d love your feedback and to connect.

Short walkthrough demo

Data Science

LightweightMMM Attribution Pipeline

Built a LightweightMMM pipeline leveraging Meta's Robyn dataset plus Google Ads and Analytics exports to estimate channel elasticities, adstock lags, and seasonality and produce posterior predictive estimates and marginal ROI curves.

>Meta's Robyn dataset integration with Google Ads and Analytics
>Exported parameterized scenario grids and hero mixes for interactive Tableau decisioning
>Delivered pilot reallocation with estimated 9% incremental revenue uplift

LightweightMMMJAXPythonSQLTableau

Data Science

TransferEconomics

Engineered a log-linear OLS valuation model on an independently compiled panel of 8,490 player-seasons from Europe's top 5 leagues, constructing a composite performance index and controlling for wages, age, position and prior value to isolate transfer timing and origin effects.

>Log-linear OLS valuation model on 8,490 player-seasons panel
>Measured significant mid-season valuation penalty and homegrown discount
>Produced ranked undervaluation lists and interactive Tableau scouting dashboard

PythonPandasStatsmodelsTableau

AI/ML

CharGen | 3D Character Generation

AI-powered 3D character synthesis from text descriptions using dual stable diffusion pipelines, DECA model for facial processing, and FReMP framework deployment. Published research paper on text-to-3D synthesis techniques.

>Dual stable diffusion pipelines for face and body generation
>DECA model integration for realistic mesh and texture
>Published academic paper on synthesis methodology

PythonAI/MLFReMPDECA

Web Dev

IPL.com | Cricket Platform

Multi-page web application for Indian Premier League with comprehensive statistics, FAQ system, ticket booking functionality, and personalized team selection features.

>Interactive statistics dashboard
>Ticket booking system integration
>Personalized team selection algorithm

ReactJavaScriptUI/UX

Data Science

Optimal IPL Team Selection

Predictive ML model for optimal playing 11 selection using supervised learning techniques, analyzing player statistics, match performance, and historical data for data-driven team composition.

>Supervised learning model implementation
>Player performance analysis engine
>Data-driven team optimization

PythonMLData Analysis

Systems

Hadoop MapReduce Implementation

Custom implementation of Hadoop's MapReduce framework with multi-node configuration for parallel job execution using Python and socket programming.

>Multi-node parallel processing
>Custom MapReduce implementation
>Efficient socket-based communication

PythonDistributed SystemsSocket Programming

Computer Vision

Image Noise Removal Filters

Digital image processing application implementing various noise reduction filters with adjustable parameters for exploring filter behavior and performance.

>Multiple filter implementations (median, Gaussian, etc.)
>Interactive noise level adjustment
>Real-time filter comparison

MatlabImage Processing

Analytics

Blue Bikes Analytics Dashboard

Interactive Tableau dashboard analyzing bike-sharing patterns, seasonal trends, and station popularity with geospatial visualizations for operational optimization.