Adam Rangwala

Hi There! I'm Adam Rangwala

Ex-Accenture | Software Engineer | STEM Educator | Data Analytics |
AI & LLMs

About Me

Who Am I

I’m a software engineer, data analyst, former Accenture consultant and STEM educator! My passion lies in taking data to provide unique insights.

I have worn many hats over the years as a digital consultant working with different visual and education applications, a founder of a drone imaging company and working EdTech. I focus on building creative solutions especially in the education space!

Skills

Python

Tensorflow

HTML

JavaScript

Matlab

SQL

PostgreSQL

MongoDB

PowerBI

Pandas

Git

Streamlit

Blender

OpenCV

Monday.com

Recent Projects

Explore my recent projects showcasing my skills in computer vision, remote sensing and machine learning applications

ACT Math Sprint Website

Designed, created and currently maintaining an ACT Math Practice website to give free access to practice problems, helpful ACT tips, and learning videos in a Duolingo-style learning platform. [React + Firebase with an Express API layer]

3D Reconstruction of Landmarks

Aiming to preserve the 1886 built Bissell Water Tower, I created an accurate 3D-printable model of the tower using images from drone photography, WEBODM for photogrammetry, Blender to scale, clean and polish the 3D object for 3D-printing.

OCR Pipeline for Old City Directories

This OCR pipeline extracts structured resident data from 1900 Minneapolis city directories for House Novel startup, achieving 95%+ accuracy using computer vision, Tesseract OCR, and domain-specific text processing to convert historical documents into standardized JSON for genealogical research. Built with Python, OpenCV, and comprehensive testing, it demonstrates production-ready engineering with batch processing, checkpoints, and validation systems.

Application Job Tracker

Designed and developed an automated Job Application Tracker using Google Apps Script that scans Gmail for job-related emails, intelligently extracts company names and job titles, and categorizes applications by status (Applied, Interview, Assessment, Offer, Rejected), creating a dynamic dashboard with visualizations and analytics to optimize job search efficiency and organization.

AI-Generated Image Identification

A Streamlit web application to classify images as either human or AI-generated. This is a visualization of the effectiveness of CNN models trained for a binary classification Kaggle competition.

Population Density Mapping

Population Density Mapping is a Streamlit webapp and Python package to create beautiful population density maps from the WorldPop dataset.

Face Detection Playground

Face Detection can have higher effectiveness with some pre-processing techniques like blurring and exposure correction. This Streamlit app allows you to apply a few imaging techniques to aim for a higher than 90% confidence threshold on face detection.

Image Restoration

A Streamlit web application to restore old images using classic Computer Vision techniques: median blur, bilateral filtering, and Navier-Stokes/Telea In-painting.