Om Balgude — Full-Stack Engineer specializing in Real-Time Systems and AI Products
I’m Om Balgude a full-stack engineer focused on building scalable web applications, real-time systems, AI-powered platforms, and modern digital experiences with strong architecture and clean engineering.
My work spans across backend systems, APIs, databases, real-time communication, and frontend engineering. I enjoy building end-to-end products from designing scalable backend architectures and optimizing data flow to crafting responsive, intuitive interfaces with smooth user experiences and thoughtful interactions.
I’ve worked on projects involving real-time collaboration systems, AI-assisted platforms, secure transaction workflows, browser-based code execution environments, and intelligent automation systems. My experience includes working with WebSocketss, Redis Pub/Sub, event sourcing, Docker, monorepo architectures, OCR pipelines, blockchain-integrated verification systems, JWT authentication, and scalable synchronization infrastructure.
I’m especially interested in distributed systems, real-time architectures, AI-powered products, developer tooling, performance optimization, and product-focused engineering. I care deeply about scalability, maintainability, reliability, and building applications that combine strong engineering with seamless user experiences, polished UI/UX, and thoughtful micro-interactions.
Beyond coding, I think in terms of ownership, product impact, system design, and long-term engineering decisions. I’m constantly exploring, learning, experimenting, and pushing myself to build production-ready systems that combine strong backend architecture with polished user experiences.
Trainee Software Engineer (Full Stack + AI/ML) | @Sumago Infotech Pvt. Ltd
- •Contributing to full-stack development and AI-integrated workflows for real-world business applications with a focus on scalability, maintainability, and production-ready architecture.
- •Building and improving backend APIs, database workflows, authentication systems, and frontend features using Node.js, Express.js, React, and modern web technologies.
- •Worked on AI-powered platforms including a soil analysis recommendation system, SOP-based AI guidance assistant, and intelligent workflow automation systems for improved user guidance and operational efficiency.
- •Collaborating across frontend, backend, and AI/ML teams to deliver scalable features, optimize development workflows, and improve overall product experience and system reliability.
Technologies Used
Full Stack Developer Intern | @IonCure Tech Pvt. Ltd
- •Contributed to research, planning, and development workflows for the AI Vidya Portal, helping shape scalable frontend and backend implementation strategies.
- •Developed and optimized RESTful APIs, backend workflows, and database schemas for scalable application modules and efficient data handling.
- •Worked across frontend and backend systems using React, Node.js, and Express.js to deliver responsive, production-ready features and improve overall application performance.
Technologies Used
Secure browser-based code execution platform powered by isolated Docker sandboxing.
A full-stack online coding playground enabling real-time execution of Python, Rust, and JavaScript code inside isolated Docker containers with secure sandboxing and resource-limited execution.
- •Built a browser-based code execution platform with Monaco Editor integration, multi-language execution support, and real-time output rendering.
- •Designed isolated Docker-based execution pipelines with CPU, memory, timeout, and network restrictions for secure sandboxed code execution.
- •Implemented scalable backend execution workflows with temporary containerized environments and automated cleanup mechanisms.
- •Designing secure containerized execution workflows while preventing unrestricted resource usage and unsafe code execution.
- •Managing isolated execution environments, execution timeouts, temporary file systems, and reliable stdout/stderr handling.
Technologies Used
Scalable collaborative canvas platform with real-time synchronization and event-driven architecture.
A production-oriented collaborative whiteboard platform focused on low-latency synchronization, scalable real-time communication, persistent event storage, and seamless multi-user collaboration.
- •Architected a multi-server real-time collaboration platform using Next.js, Express.js, WebSockets, PostgreSQL, and Redis Pub/Sub for scalable synchronization and communication.
- •Implemented low-latency room-based collaboration with shared canvas synchronization, persistent event storage, local caching, and real-time broadcasting.
- •Designed scalable backend workflows, JWT-based authentication, event-driven synchronization infrastructure, and monorepo-based architecture for maintainability and extensibility.
- •Maintaining consistent shared state across concurrent users while keeping synchronization latency low and interactions seamless.
- •Designing scalable real-time communication flows and synchronization logic capable of handling large concurrent traffic efficiently.
Technologies Used
Secure digital payment platform focused on transaction consistency and concurrency-safe financial workflows.
A full-stack payment platform designed around secure authentication, transactional reliability, rollback guarantees, and consistent financial state management under concurrent transfer conditions.
- •Engineered secure transaction APIs using MongoDB multi-document transactions to maintain reliable financial consistency during concurrent transfers.
- •Implemented JWT-based authentication, protected payment workflows, and responsive frontend interfaces for seamless transaction experiences.
- •Designed backend transaction flows focused on rollback guarantees, concurrency safety, and reliable state management under failure scenarios.
- •Ensuring concurrency safety and transactional consistency during simultaneous transfers and retry conditions.
- •Maintaining accurate financial state and rollback reliability under concurrent operations and edge-case failures.
Technologies Used
AI-assisted certificate verification platform with OCR analysis, tamper detection, and blockchain-backed trust workflows.
A multi-service certificate authenticity validation platform combining OCR pipelines, heuristic tamper detection, structured hashing, role-based trust management, audit logging, and optional blockchain verification.
- •Developed a full-stack certificate validation platform integrating OCR extraction, AI-assisted verification workflows, audit logging, and institution-scoped access control.
- •Built document normalization and hashing pipelines for semantic certificate verification and deterministic trust validation.
- •Integrated real-time verification updates, tamper-analysis workflows, and optional blockchain-backed document anchoring for immutable verification support.
- •Designing explainable verification workflows that combined OCR confidence, tamper analysis, anomaly scoring, and structured database matching.
- •Managing trust boundaries, role-based access control, and multi-service communication across frontend, backend, AI, and blockchain layers.
Technologies Used
CNN-based image classification model for Fashion MNIST dataset using deep learning workflows.
A convolutional neural network model trained on the Fashion MNIST dataset for clothing image classification with optimized preprocessing, model training, and evaluation workflows.
- •Developed and trained a convolutional neural network for multi-class fashion image classification using TensorFlow and Keras.
- •Implemented preprocessing, normalization, model training, evaluation, and prediction workflows for image-based deep learning tasks.
- •Analyzed model accuracy, loss metrics, and prediction performance across multiple clothing categories.
- •Improving classification accuracy while reducing overfitting during model training.
- •Optimizing CNN architecture and preprocessing workflows for better prediction performance.
Technologies Used
Deep learning-based housing price prediction model using regression analysis.
A regression-focused machine learning project that predicts housing prices using neural networks, feature preprocessing, and supervised learning workflows.
- •Built and trained a deep neural network model for predicting housing prices using structured numerical datasets.
- •Implemented data preprocessing, normalization, training-validation workflows, and regression-based evaluation metrics.
- •Analyzed prediction accuracy and optimized training performance using loss monitoring and model tuning techniques.
- •Handling feature preprocessing and improving prediction accuracy for continuous-value regression outputs.
- •Balancing model complexity and generalization performance during neural network training.
Technologies Used
Deep learning-based sentiment analysis system for classifying movie reviews.
A natural language processing project focused on sentiment classification of movie reviews using neural networks, text preprocessing, and binary classification workflows.
- •Developed a neural network-based sentiment analysis model for classifying movie reviews into positive and negative categories.
- •Implemented text preprocessing, tokenization, vectorization, and training workflows for NLP-based classification tasks.
- •Evaluated model accuracy and optimized classification performance using deep learning techniques.
- •Handling text preprocessing and feature representation for accurate sentiment classification.
- •Improving model generalization and reducing classification errors on unseen reviews.
Technologies Used
Open to full-time opportunities, freelance projects, and collaborations focused on scalable products, AI-powered applications, and modern web experiences.