Data Analyst | SQL, Python, Power BI, Tableau | Data Visualization & Business Intelligence | B.Sc. Data Science 2026

github

Ajit Mahapatro  • Peerlist

Developed QuantAlpha, an AI-powered stock market analytics web application, as part of a team project during the Machine Learning with Python Internship by Adhco Network. Built with React, TypeScript, FastAPI, and Python, the project provides portfolio analytics, backtesting, stock research signals, and market trend insights using yfinance data. Worked on integrating the frontend and backend, improving dashboard usability, handling real-world data issues, and deploying the application for practical use. Mentored by Ganesh Sir.

QuantAlpha

Developed a high-fidelity, two-page Power BI intelligence suite to analyze $899.90M in US retail revenue. The project transitioned raw transactional data into an interactive executive tool focused on identifying regional profit leakage and optimizing product mix. Key Contributions: AI-Driven Root Cause Identification: Implemented a Decomposition Tree to allow stakeholders to drill down from total sales ($899M) into specific regions, retailers, and product categories to find performance drivers. Advanced Spatial Analytics: Leveraged Bubble Maps to visualize market density across US cities, identifying top-performing states and regional sales clusters. Custom Financial Engineering: Authored complex DAX measures for Operating Margin %, Profit per Unit, and Sales Velocity to assess merchandising efficiency at a granular level. Glass UI/UX Design: Engineered a dark-themed, app-style interface featuring synced cross-page navigation and conditional formatting (Data Bars) to reduce cognitive load for executive users. Technical Stack: Power BI Desktop, DAX (Data Analysis Expressions), Data Modeling, Advanced Visualization, Retail Strategy Analysis.

Adidas Global Sales & Profitability Diagnosti

• Decoupled Architecture Deployment: Architected and deployed a production-style, decoupled software platform separating a high-speed asynchronous processing backend from lightweight browser execution assets. • Asynchronous API Infrastructure: Engineered an asynchronous FastAPI microservice backend framework utilizing strict Pydantic schemas for request/response data validation and dynamic environment configuration. • Low-Latency Model Inference: Integrated optimized Llama-3.3 model inference pipelines via Groq API to ingest structural database profiles and output deterministic, decision-ready performance analytics. • Real-Time Statistical Visuals: Developed an interactive data profiling and visualization module using client-side Plotly.js to render real-time multi-variate statistical distributions, trends, and correlation matrices. • Automated Data Extraction: Programmed an automated web scraping component utilizing BeautifulSoup4 to programmatically extract, clean, and structure raw text arrays from targeted web pages for pipeline processing.

Stark Data Science Assistant – Full-Stack AI 

Developed an end-to-end HR analytics solution combining machine learning predictions with an interactive Power BI dashboard to analyze employee attrition patterns and retention risks. Performed exploratory data analysis (EDA), data preprocessing, and feature engineering using Python, Pandas, and SQL. Built a Random Forest classification model to predict employee attrition risk and integrated analytical insights into a multi-page Power BI dashboard. The dashboard included: • attrition trend analysis • workforce demographics • predictive risk segmentation • AI-driven root cause analysis • KPI tracking and retention insights Designed an application-style dashboard experience using custom Power BI UI components, interactive tooltips, decomposition trees, and business storytelling visuals. Key business insights generated: • Estimated $11.85M in potential attrition-related salary impact • Identified high-risk employee segments through predictive analytics • Projected retention improvement opportunities worth approximately $474K Tech Stack: Python, SQL, Power BI, Pandas, Scikit-Learn, Random Forest

Employee Attrition Risk & Retention Analytics

Developed an interactive Power BI analytics dashboard using the Adidas US Retail dataset to analyze sales performance, regional trends, product profitability, and operational KPIs. Performed data cleaning and transformation directly in Power BI using Power Query and built DAX measures for KPI tracking, profit analysis, trend comparison, and retailer performance analysis. Key dashboard features included: • executive KPI reporting • decomposition tree-based root cause analysis • geographic sales visualization using bubble maps • product-level profitability analysis • interactive drill-down reporting The dashboard was designed with a modern business-intelligence reporting approach focused on usability, storytelling, and executive decision support. Business insights generated: • Identified top-performing retailer and regional sales segments • Analyzed operating margin and product profitability trends • Visualized high-growth sales clusters across multiple regions Tech Stack: Power BI, DAX, Power Query, Data Visualization, Business Intelligence

Adidas Retail Strategy & Sales Analytics Dash

Led a team in building an end-to-end Student Placement Prediction Web App that uses Machine Learning to estimate whether a student is likely to be placed based on factors such as CGPA, internships, aptitude, communication skills, projects, branch, and backlogs. The project combines an XGBoost-based prediction pipeline with a FastAPI backend and a responsive web interface to deliver real-time predictions along with confidence scores. I also helped drive the deployment-ready structure by integrating session handling, reset functionality, and a retraining workflow for updating the model. This project strengthened my experience in machine learning, backend development, and building practical web-based predictive systems.

Student Placement Prediction Web App

Collaborated in a team-based analytics project focused on identifying operational factors and hidden patterns contributing to airline delays. Worked primarily on data preprocessing, cleaning, exploratory data analysis (EDA), and visualization workflows using Python and Pandas. Analyzed large-scale flight operations datasets to identify trends across airlines, airports, routes, and delay categories. Project contributions included: • data preprocessing and transformation • delay trend analysis • route and carrier-based visualizations • operational pattern discovery using EDA • supporting association rule mining workflows using FP-Growth The project helped uncover hidden relationships between routes, airports, carriers, and delays to support operational insights and resource planning decisions. Tech Stack: Python, Pandas, EDA, Data Visualization, Statistical Analysis, FP-Growth

Airline Delay Pattern Discovery ✈️

Built an end-to-end predictive retention system combining a Random Forest classifier (93% accuracy) with a custom Power BI dashboard. Quantified $11.85M in attrition risk and designed a dynamic 'Root Cause Matrix' to identify high-churn cohorts, enabling targeted retention strategies and projected savings of $474K.

Attrition Analysis: Integrating Random Forest

Engineered an Unsupervised Learning engine (FP-Growth) to analyze large-scale aviation logs, detecting hidden operational anomalies without labeled data. Defined "High Lift" logic to isolate specific carrier-route combinations with a 90% probability of disruption. Visualized systemic bottlenecks and delay propagation using Graph Theory (NetworkX). Tech Stack: Python, Pandas, MLxtend, NetworkX.

Airline Delay Pattern Discovery

• Star Schema Warehouse Modeling: Engineered a relational Star Schema data warehouse architecture, decoupling system attributes into structured dimension tables (dim_clusters, dim_nodes) linked to a central time-series fact table. • Telemetry Pipeline Automation: Built a vectorized Python pipeline utilizing C-optimized Pandas and NumPy arrays to generate 4,608 synchronized system logs tracking 64 Lakh+ individual transactions across international banking nodes. • Statistical Anomaly Injection: Coded an exponential random distribution algorithm to establish healthy latency benchmarks, while programmatically engineering an acute 90-minute network degradation simulation spiking lag to 3,500 ms. • Advanced Visual Diagnostics: Deployed Tableau Fixed Level of Detail (LOD) expressions and conditional matrix calculations to calculate the structural "blast radius" of infrastructure failures and isolate regional SLA breaches.

Global Enterprise Ledger & Database Replicati

⚛️ Project Deep Dive: QOPS (Quantum Operational Intelligence System) Built an award-winning project from the Amaravati Quantum Valley Hackathon (Problem Statement: AQVH915). QOPS was engineered to solve the "black box" nature of quantum execution by providing a transparent, high-fidelity management layer between researchers and hardware. The entire system is built into two powerful modules: 1. Quantum Job Monitoring Dashboard (Intelligence Module) This module acts as the "eyes" of the system, using a MERN stack and custom REST APIs to provide real-time telemetry. **Real-time Analytics**: Visibility into job status, backend load, and queue analytics for @IBM Quantum backends (like ibm_torino and ibm_marrakesh). **Backend Health Monitoring**: Live tracking of system uptime and error rates to suggest the most efficient hardware for a circuit. **Omnichannel Access**: Includes a native Mobile App version so researchers can track 48+ hour quantum jobs on the go. 2. Secure Quantum Execution Engine (Execution Module) The "engine" of the system, powered by Python and @Qiskit, handling the lifecycle of quantum code from design to submission. **Visual Circuit Architect**: A pro-grade Drag-and-Drop interface for designing quantum gates (H, CX, RX, etc.). **Automated Hardware Translation**: Real-time generation and validation of OpenQASM 3.0 code from visual designs. **DevOps Command Center**: A specialized dashboard to track Total, Completed, Running, and Pending jobs with full execution logs. **Enterprise RBAC**: A secure, dual-role system. While Users have a private view restricted to their own submissions, the Admin Dashboard provides a global overview of all jobs across the organization—ensuring full operational control. 💻 The Tech Stack: Full-Stack: MongoDB, Express, React, Node.js (MERN) Quantum core: Python, Qiskit, IBM Quantum API Bridge: Custom REST API for job polling and telemetry data Hardware Layer: OpenQASM 3.0

Qops

SQL-driven analysis of pizza sales data, extracting key insights on orders, revenue, popular pizzas, and operational patterns to inform business strategies.

SQL Project: Pizza Sales Analysis

This project focuses on analyzing user behavior data collected from various mobile devices. The dataset includes key metrics such as app usage time, screen on time, battery drain, and user demographics. The primary objective of this analysis is to uncover insights into user behavior patterns and trends, which can inform app development and optimization strategies.

Trend Analysis With Phone Data

Led a 7-member team in developing a full-stack financial analytics platform focused on stock market analysis, portfolio intelligence, and machine learning-driven forecasting. • Developed data processing and feature engineering pipelines using Python, Pandas, and Scikit-Learn for financial market analysis. • Contributed to FastAPI-based REST API development for serving analytics, forecasting outputs, and portfolio insights. • Worked on frontend-backend integration workflows to connect analytics services with interactive dashboards. • Implemented modules for portfolio performance tracking, backtesting, risk analysis, volatility monitoring, and market research signals. • Addressed real-world engineering challenges including API rate limits, unreliable third-party market data sources, deployment stability, and system reliability. • Collaborated within a 7-member team using modular development practices, feature ownership, testing, and iterative product improvements. Key Features: • Portfolio Performance Tracking • Risk & Volatility Analytics • Sharpe Ratio Monitoring • Market Trend Analysis • Backtesting Engine • Research Signal Dashboards Tech Stack: Python, FastAPI, React, Scikit-Learn, XGBoost, Financial Analytics

QuantAlpha – Financial Analytics & Market Pre

Collaborated in a 6-member team during the AP Amaravati Quantum Valley Hackathon to develop a monitoring and operational intelligence platform for IBM Quantum job execution workflows. • Contributed to the design and development of a system for monitoring quantum job execution, backend activity, and queue status across IBM Quantum platforms. • Worked with Qiskit and IBM Quantum APIs to support execution tracking, job lifecycle monitoring, and backend status analysis. • Assisted in developing analytics and telemetry workflows for processing execution-related insights and operational metrics. • Contributed to dashboard visualization features for monitoring queue activity, execution progress, backend availability, and system performance. • Participated in API integration, testing, debugging, and collaborative feature development within a fast-paced hackathon environment. • Collaborated on system architecture discussions and operational monitoring concepts to improve visibility into quantum computing workloads. 🏆 Secured 2nd Prize at the AP Amaravati Quantum Valley Hackathon under the Quantum Job Tracker challenge. Tech Stack: Python, Qiskit, IBM Quantum APIs, REST APIs, System Monitoring, Analytics Dashboards