Hello World, I am

Heramb Devbhankar

A software engineer with a Master's degree from the University of Southern California and over 6 years of experience. I specialize in both front-end and back-end development, with expertise in Algorithmic Programming and Machine Learning. I primarily use Python, React, and Swift. I build dynamic Web and iOS apps focusing on seamless user experiences. Passionate about exploring new technologies, I continually advance my skills in Software Engineering and Machine Learning.

Available for new projects and Full-time jobs

Technology Stack

Professional Experience

Senior Software Engineer

Apollo Global Management

Nov 2024 - Present

Designed and developed a large-scale risk data management platform for Apollo’s investment and risk teams, enabling efficient retrieval and processing of over ~70,000 rows and 350 columns (≈465 MB) of financial data through scalable APIs and automated workflows. Built and deployed RESTful APIs using Python (FastAPI), optimizing data access with complex SQL Server stored procedures and improving performance with Protobuf serialization and Redis caching for high-volume data visualization. Engineered and orchestrated Temporal workflows to automate data ingestion and risk dataset generation, integrating multiple data sources such as quant models, monthly positions, and risk metrics into robust SQL Server pipelines with polling, scheduling, and fault-tolerant orchestration. Delivered a high-performance automated data pipeline that reduced processing and API response times by over 60%, improved reliability, and became a foundational component of Apollo’s portfolio risk analytics platform.

Software Engineer

SAF-AFTRA Foundation

May 2023 - Oct 2024

Led backend development using Python-Django, optimizing data management with Django ORM and implementing email systems to enhance functionality. Streamlined integration with dynamic frontends (iOS and ReactJS), automated deployment pipelines, and implemented scalable solutions to ensure high availability under variable traffic conditions with.

Software Engineer

Eye Labs Inc

Sep 2022 - Apr 2023

Developed an iOS application in SwiftUI utilizing the Vision framework and machine learning models to detect eye anomalies using the front camera. Integrated advanced UIKit features, implemented mathematical algorithms for facial and eye direction analysis, and established a baseline product for FDA clinical trials, successfully tested on eye patients.

Research Assistant | Master of Science in Computer Science

University of Southern California

2021 - 2022

Pursued a Master’s in Computer Science at USC, focusing on advanced topics like AI, data science, and software engineering, while gaining hands-on experience through projects and research assistanship. | Developed models for tracking Drosophila fly movement and gene expression using LSTM and CNN, and researched face recognition systems trained on synthetic data with StarGAN for data augmentation and multi-domain image translations.

Software Engineer - Machine Learning

IOTric

May 2019 - Dec 2021

Built a web application that extracts text from distorted invoice and receipt images using a three-stage pipeline (object detection, text detection, and extraction). Deployed the object detection model on AWS EC2 with a REST API and optimized the app to achieve a 100ms response time per document for efficient text extraction.

Notable Projects

View Source

Heart MRI Segmentation

Source

This heart segmentation project uses a U-Net model in PyTorch with Dice loss for segmenting heart structures in MRI images. Hyperparameter tuning (e.g., learning rates, batch sizes) optimizes model performance, aiming for high segmentation accuracy.

Visit Demo Colab Notebook...

Python

OpenCV

PyTorch

Computer Vision

Deep Learning

StockSearch WebApp

Private source

A full-stack web application with ReactJS for the front-end, allowing users to search stocks and view analytics through Highcharts. The app functions like a personalized version of Robinhood, enabling real-time stock tracking and data visualization.

DEMO VIDEO

Python

Django

ReactJS

MySQL

AWS RDS

AWS EC2

StockSearch iOS App

Private source

A stock search app built with SwiftUI for iOS, enabling users to search stocks, view detailed analytics, and track real-time market data. The app offers a personalized experience, similar to Robinhood, with smooth and interactive visualizations to help users make informed decisions.

DEMO VIDEO

Swift

SwiftUI

iOS App Development

UIkit

MVC

View Source

pip install spoken2written

Source

Spoken2Written is an open-source Python package designed to convert spoken text into its corresponding written form, capturing nuances like informal speech patterns, contractions, and verbal fillers, making it suitable for transcription and text processing applications.

Explore the open source package...

Python

NLTK

Spacy

Natural Language Processing

Open-Source

Skills

Machine Learning, AI and Data Science

I develop AI solutions using Python, PyTorch, and TensorFlow, specializing in machine learning, data science, computer vision, and NLP.

Backend, API, and Databases

I deliver secure, efficient, and scalable backend REST API solutions using Python, Django, and Flask, ensuring seamless integration and robust performance for modern applications.

Web App Development

I specialize in providing functional web services that include custom website app development, WordPress development, e-commerce solutions, and more.

Mobile App

I design and develop stunning, feature-rich apps for iOS, Android using SwiftUI and React Native, turning your vision into reality with exceptional quality and performance.

AWS and Cloud Services

Building scalable cloud architectures and AWS solutions using Infrastructure as Code (IaC) with Terraform and CloudFormation, ensuring automation, reliability, and seamless deployment.

Data Engineering and Analytics

I design and build scalable data pipelines and analytics solutions using tools like Python, SQL, and Apache Spark, transforming raw data into actionable insights with efficiency and precision.