Portfolio Projects

Production-grade applications at scale, across .NET, cloud-native backends, real-time systems, IoT, AI/ML, 3D graphics, and embedded systems — spanning aerospace, fintech, healthcare, automotive, and SaaS

Fleet Flows - Real-Time Vehicle Telematics Dashboard

Fleet Flows - Real-Time Vehicle Telematics Dashboard

Enterprise fleet management platform, live tracking and predictive maintenance

Built an enterprise telematics dashboard processing real-time CAN bus data from 500+ vehicles simultaneously. Implemented WebSocket architecture for sub-200ms latency updates, predictive maintenance algorithms analyzing fuel efficiency curves and brake wear patterns, and driver behavior scoring system.

ReactASP.NET Core 10SignalREF Core 10+5 more
View Details

Key Features:

  • Real-time vehicle tracking with map clustering
  • Predictive maintenance alerts (85% accuracy)
  • Custom metrics: fuel efficiency, engine load, brake wear
  • Event-driven microservices architecture
  • Time-series data optimization (TimescaleDB)

Architecture:

ASP.NET Core 10 Minimal APIs event-driven microservices, SignalR Core with Redis Backplane for real-time, EF Core 10 with PostgreSQL/TimescaleDB, AWS IoT Core simulation layer.

Full Tech Stack:

ReactASP.NET Core 10SignalREF Core 10WebSocketsAWS IoT CorePostgreSQLRedisDocker
Government VR Learning Platform

Government VR Learning Platform

Multiplayer VR system for 2000+ students at World Skill Centre

Sole developer for government VR/multiplayer learning platform deployed at World Skill Centre, Odisha. Architected real-time collaboration system supporting 50+ concurrent users per session with <200ms latency. Handled complete project lifecycle from requirements to deployment.

C#VRASP.NET Core / Custom UDPAWS+1 more
View Details

Key Features:

  • 50+ concurrent users per VR session
  • <200ms collaboration latency
  • Cross-platform (PC, VR, Mobile)
  • Real-time state synchronization
  • Custom VR interaction framework

Architecture:

Custom UDP networking stack in C# with server-authoritative state sync, ASP.NET Core backend for user management and session orchestration, AWS backend for user management, custom VR framework for educational interactions, optimized for standalone VR headsets.

Full Tech Stack:

C#VRASP.NET Core / Custom UDPAWSWebSockets
Enterprise AR/IoT System - TATA Aerospace

Enterprise AR/IoT System - TATA Aerospace

Real-time defect detection for aerospace maintenance

Developed real-time defect detection system for aerospace MRO (Maintenance, Repair, Operations) achieving <100ms latency for video processing and AR overlay. Built in Unity (C#) with .NET Aspire cloud orchestration, complete solution deployable for 2 major airlines with TATA Aerospace.

HoloLens 2AzureC# / UnityASP.NET Core+6 more
View Details

Key Features:

  • <100ms latency for real-time processing
  • HoloLens 2 AR client
  • Python-based AI defect detection
  • Azure cloud infrastructure
  • React management portal

Architecture:

HoloLens 2 client with spatial mapping, .NET Aspire orchestration layer managing Azure Functions lifecycle, C# Unity client on HoloLens 2, ASP.NET Core management API, Python computer vision models, WebRTC for video streaming, CosmosDB for defect records, SignalR for real-time updates.

Full Tech Stack:

HoloLens 2AzureC# / UnityASP.NET Core.NET AspirePythonComputer VisionWebRTCReactCosmosDB
MediScan AI - Medical Imaging Collaboration Platform

MediScan AI - Medical Imaging Collaboration Platform

HIPAA-compliant radiology annotation platform with AI-assisted region detection

Developed a medical imaging SaaS platform for radiologists with AI-powered segmentation using pre-trained U-Net models. Implemented real-time collaborative annotation supporting 10+ concurrent radiologists per case with conflict resolution.

ReactPythonFastAPIPyTorch+4 more
View Details

Key Features:

  • DICOM viewer with AI-assisted ROI detection
  • Real-time collaborative annotations (WebRTC)
  • HIPAA-compliant data handling architecture
  • ML model serving via FastAPI
  • Annotation version control

Architecture:

FastAPI microservices, PyTorch model serving with TorchServe, WebRTC for real-time collaboration, encrypted S3 storage with pre-signed URLs, audit logging for compliance.

Full Tech Stack:

ReactPythonFastAPIPyTorchWebRTCPostgreSQLDockerAWS S3
StrategyLab - Algorithmic Trading Backtester

StrategyLab - Algorithmic Trading Backtester

High-performance trading strategy backtester with C++ compute engine

Built a fintech platform for backtesting algorithmic trading strategies with C++ compute engine achieving 10x faster execution than pure Python. Visual flow builder for strategy composition with real-time P&L visualization.

ReactPythonC++WebAssembly+3 more
View Details

Key Features:

  • Visual strategy builder (no-code interface)
  • C++ compute engine (compiled to WebAssembly)
  • Historical data pipeline (Yahoo Finance API)
  • Financial metrics: Sharpe ratio, drawdown, win rate
  • Strategy comparison dashboard

Architecture:

C++ core logic exposed via WebAssembly for browser execution, Python data pipeline with Pandas/NumPy, React Flow for visual strategy builder, PostgreSQL for strategy storage.

Full Tech Stack:

ReactPythonC++WebAssemblyPandasPostgreSQLDocker
BuildPulse - Distributed CI/CD Monitor

BuildPulse - Distributed CI/CD Monitor

Real-time monitoring dashboard for distributed build pipelines

Developed a DevOps monitoring platform tracking 50+ concurrent build jobs across distributed infrastructure. Real-time log aggregation, failure analysis with automatic retry logic, and resource utilization tracking.

ReactASP.NET CoreDockerRedis+3 more
View Details

Key Features:

  • Real-time build pipeline visualization
  • Distributed job queue management
  • Failure pattern detection & auto-retry
  • Resource utilization tracking (CPU/Memory/Disk)
  • Integration with Jenkins/GitLab CI

Architecture:

ASP.NET Core SignalR for real-time pipeline updates, background job workers via .NET hosted services for job simulation, Redis for job queue, WebSocket real-time updates, Docker containerization for distributed execution, Prometheus metrics collection.

Full Tech Stack:

ReactASP.NET CoreDockerRedisWebSocketsPrometheusGrafana
TenantCore - Multi-Tenant SaaS Foundation

TenantCore - Multi-Tenant SaaS Foundation

Production-ready SaaS boilerplate with enterprise-grade tenant isolation

Architected a multi-tenant SaaS platform using .NET 10 Clean Architecture (Onion/DDD) and C# 14 with schema-per-tenant isolation, role-based access control (RBAC), and subscription management. Built for horizontal scalability with complete audit logging.

.NET 10ASP.NET CoreEF Core 10React+5 more
View Details

Key Features:

  • Multi-tenancy with schema-per-tenant isolation
  • Complete RBAC system with custom roles
  • Stripe subscription management + webhooks
  • Comprehensive audit logging
  • Tenant-specific data encryption

Architecture:

PostgreSQL with schema-per-tenant isolation, Redis for session management, .NET 10 Clean Architecture (Onion/DDD), EF Core 10 Interceptors for tenant-context switching, ASP.NET Core claims-based RBAC for auth, webhook processing for payment events, Kubernetes deployment with per-tenant resource limits.

Full Tech Stack:

.NET 10ASP.NET CoreEF Core 10ReactPostgreSQLRedisDockerKubernetesStripe
Venturi MyCar - Automotive IoT Platform

Venturi MyCar - Automotive IoT Platform

IoT platform converting regular vehicles into connected smart cars

Developed an automotive IoT platform enabling vehicle connectivity through custom sensor installations. Real-time monitoring of vehicle metrics (location, speed, RPM, gear, orientation), health diagnostics (CEL, ABS, oil/coolant status), and AI voice control integration. Transformed a 2008 Honda Accord into a fully connected IoT vehicle with CAN bus integration and cloud telemetry.

IoTCAN BusThree.jsAWS+2 more
View Details

Key Features:

  • Real-time vehicle telemetry (speed, RPM, gear, yaw/pitch/roll)
  • Live GPS tracking (parked/moving status)
  • Vehicle health monitoring (CEL, ABS, oil, coolant)
  • Custom digital gauge cluster with AI voice commands
  • CAN bus integration for vehicle control (AC, windows)
  • 3D vehicle visualization (Three.js embedded console)
  • Tiered subscription access management
  • IoT device provisioning and management

Architecture:

IoT sensors with cellular connectivity, CAN bus protocol integration, cloud backend for telemetry storage, Three.js 3D visualization, subscription management system, real-time data streaming.

Full Tech Stack:

IoTCAN BusThree.jsAWSNode.jsC++
CodeSync - Real-Time Collaborative Code Editor

CodeSync - Real-Time Collaborative Code Editor

Google Docs for code with real-time collaboration and sandbox execution

Developed a real-time collaborative code editor supporting 10+ simultaneous users per session. Implemented Operational Transformation for conflict-free editing, integrated Monaco editor (VS Code engine), and sandboxed code execution.

ReactASP.NET CoreMonaco EditorWebRTC+2 more
View Details

Key Features:

  • Real-time multi-user editing with cursor positions
  • Conflict-free synchronization (Operational Transformation)
  • Syntax highlighting for 20+ languages
  • Sandboxed code execution (Docker containers)
  • Integrated chat + voice communication

Architecture:

ASP.NET Core SignalR server handling Operational Transformation, Docker-based execution sandboxes, WebRTC for peer-to-peer communication, Monaco editor integration, Docker-based code execution sandboxes with resource limits, WebSocket real-time sync.

Full Tech Stack:

ReactASP.NET CoreMonaco EditorWebRTCDockerWebSockets
FactorySense - Industrial IoT Sensor Network

FactorySense - Industrial IoT Sensor Network

Digital twin platform for industrial facilities with predictive maintenance

Built an industrial IoT platform visualizing real-time sensor data from manufacturing facilities. Implemented edge computing logic, anomaly detection algorithms, and 3D facility visualization with digital twin concept.

C++ReactThree.jsPython+3 more
View Details

Key Features:

  • Real-time 3D facility visualization (Three.js)
  • Multi-sensor support (temperature, pressure, vibration, humidity)
  • Anomaly detection with ML models
  • Predictive maintenance alerts
  • MQTT protocol for low-bandwidth communication

Architecture:

C++ edge agents with MQTT pub/sub, Python anomaly detection models, InfluxDB time-series storage, Node.js gateway for protocol translation, Three.js 3D rendering of facility state.

Full Tech Stack:

C++ReactThree.jsPythonMQTTInfluxDBNode.js
Steel Warehouse Configurator Webapp

Steel Warehouse Configurator Webapp

3D WebGL warehouse SaaS with real-time BOM generation

Built a complex 3D WebGL configurator SaaS for steel warehouse design with multi-user support and cloud storage. Features real-time building customization, drag-and-drop positioning, intelligent beam merging, opening placements, and comprehensive BOM generation. Optimized for browser constraints (2GB memory limit) using custom shaders and intelligent geometry generation.

ASP.NETUnity C#WebGLAWS S3+4 more
View Details

Key Features:

  • Multi-building design with drag-merge functionality
  • Real-time 3D visualization with WebGL optimization
  • Comprehensive customization (openings, lean-tos, roof types, colors)
  • Smart beam structure generation and merging
  • Auth0 authentication with AWS DynamoDB storage
  • Environment and time-of-day visualization
  • JSON-based building configuration export/import
  • BOM generation for procurement

Architecture:

S3-hosted static WebGL build, DynamoDB for building data persistence, Auth0 for authentication, custom WebGL shaders for memory optimization, intelligent LOD system for performance.

Full Tech Stack:

ASP.NETUnity C#WebGLAWS S3C++DynamoDBAuth0Three.js
QueryCraft - Natural Language to SQL Converter

QueryCraft - Natural Language to SQL Converter

AI-powered SQL query generator with query optimization suggestions

Created an AI-powered database query tool converting natural language to optimized SQL. Integrated Claude API for semantic parsing, added schema understanding, join optimization suggestions, and index recommendations.

ReactPythonLangChainClaude API+2 more
View Details

Key Features:

  • Natural language to SQL conversion (90%+ accuracy)
  • Query execution with result visualization
  • Query optimization suggestions
  • Index recommendations based on query patterns
  • Schema understanding and relationship inference

Architecture:

Claude API integration via LangChain, FastAPI backend for query processing, PostgreSQL for sample databases, React data grid for results, query execution plan visualization.

Full Tech Stack:

ReactPythonLangChainClaude APIPostgreSQLFastAPI
DroneSwarm - Autonomous Multi-Agent Simulator

DroneSwarm - Autonomous Multi-Agent Simulator

Multi-drone pathfinding simulator with collision avoidance algorithms

Developed a multi-agent drone simulation platform with advanced pathfinding (A*, RRT), collision avoidance, and task allocation algorithms. Built custom physics engine and 3D visualization using WebGL.

C++ReactWebGLThree.js+1 more
View Details

Key Features:

  • Multi-drone simulation (50+ agents)
  • Pathfinding algorithms: A*, RRT, Dynamic Window Approach
  • Collision avoidance with velocity obstacles
  • Task allocation optimization
  • Real-time 3D visualization with physics

Architecture:

C++ physics engine compiled to WebAssembly, React UI with Three.js 3D rendering, multi-threading for parallel agent calculations, spatial partitioning for collision detection.

Full Tech Stack:

C++ReactWebGLThree.jsWebAssembly
3D Dental Bracket Configurator SaaS

3D Dental Bracket Configurator SaaS

AI-powered orthodontic bracket placement and jig generation webapp

Created a WebGL-based dental bracket configurator with AI tooth segmentation for orthodontists. System imports 3D jaw scans, segments individual teeth using hosted python based segmentation AI, automatically places brackets based on FDI indices, and generates support jigs for installation. Includes predictive tooth movement simulation and automated jig generation for 3D printing.

Three.jsReactNode.jsWebGL+4 more
View Details

Key Features:

  • Hosted AI tooth segmentation (15-second processing)
  • Automatic bracket placement by FDI index
  • Intelligent bracket base curvature for force distribution
  • Tooth rotation simulation (predictive view)
  • Multi-stage support jig generation (3-4 teeth per jig)
  • Boolean operations for precise jig-tooth fitting
  • STL export for 3D printing (SLA, dental resin)
  • Orthographic views and occlusal contact visualization

Architecture:

S3-hosted static WebGL application, external AI segmentation API (PyTorch-based, GPU server), AWS Lambda for serverless processing, procedural jig mesh generation with spline algorithms, manifold mesh Boolean operations.

Full Tech Stack:

Three.jsReactNode.jsWebGLC++PyTorchAWS S3Lambda
Enterprise Drystack Masonry BOM Generator SaaS

Enterprise Drystack Masonry BOM Generator SaaS

High-performance brick wall visualization with shader-based rendering

Developed a complex masonry rules-following brick wall & BOM generator processing 7M+ bricks in real-time. Shader-based rendering system generates accurate brick patterns, motion joints, cutouts, and BOM calculations. Replaces traditional mortar with plastic connectors, following strict masonry rules for realistic and safe designs.

C++Compute ShadersWebGLAzure+1 more
View Details

Key Features:

  • Shader-based rendering (7M+ bricks, zero lag)
  • Compute shader + single mesh system for performance
  • Dual view rendering (floor plan + facade)
  • Dynamic motion joints and cutout integration
  • Pitch cutout support with proper brick disruption
  • Real-time BOM generation (whole bricks, cuts, layers)
  • UIToolkit integration with hierarchy/inspector
  • Save/load design with integrity preservation

Architecture:

Compute shader-based brick generation, procedural mesh system, masonry rule engine, real-time BOM calculator, Azure-hosted WebGL build, file serialization for design persistence.

Full Tech Stack:

C++Compute ShadersWebGLAzureOpenGL
Distributed MMORTS Platform

Distributed MMORTS Platform

Real-time strategy game serving 10K+ concurrent users

Architected real-time multiplayer platform with AWS microservices backend. Implemented low-latency TCP socket communication (<100ms), server-authoritative game state synchronization, and DynamoDB for persistent world data across all players globally.

C#ASP.NET Core.NET Native AOTNode.js+4 more
View Details

Key Features:

  • 10K+ concurrent users supported
  • Server-authoritative state management
  • Real-time world map synchronization
  • AWS Lambda microservices architecture
  • DynamoDB with optimized query patterns

Architecture:

Authoritative ASP.NET Core C# game server on AWS EC2, .NET Native AOT Lambda functions for authentication and matchmaking, DynamoDB for player/world state, DynamoDB for player/world state, TCP sockets for real-time communication, OneSignal for push notifications.

Full Tech Stack:

C#ASP.NET Core.NET Native AOTNode.jsAWS LambdaDynamoDBWebSocketsTCP Sockets
AI-Powered Collaborative Chat Platform

AI-Powered Collaborative Chat Platform

Cross-platform chat with AI integration (pre-ChatGPT era)

Built cross-platform chat application with AI integration supporting document attachment, image recognition, video processing, and collaborative group chat. Native implementations for iOS (Swift) and Android (Kotlin) with identical UI and shared AWS backend.

React NativeSwiftKotlinAWS Amplify+3 more
View Details

Key Features:

  • AI-powered responses with multimodal input
  • Collaborative group chat with AI as participant
  • Native iOS (Swift) and Android (Kotlin) apps
  • Document generation within chat
  • Stripe subscription management

Architecture:

AWS Amplify for authentication, API Gateway + Lambda for backend logic, DynamoDB for chat storage, S3 for file uploads, Stripe for subscription billing, native mobile apps with shared API.

Full Tech Stack:

React NativeSwiftKotlinAWS AmplifyLambdaDynamoDBStripe
Industrial Training Visualization Platform

Industrial Training Visualization Platform

Oil rig path simulation with mathematical spline generation

Created industrial simulation platform with complex mathematical path generation using Bezier splines and tangential equations. Cross-platform desktop application for drill operator training with real-time 3D visualization of drilling paths.

C++OpenGLQtQML+1 more
View Details

Key Features:

  • Real-time Bezier spline generation
  • Complex mathematical path calculations
  • 3D visualization of drilling paths
  • Cross-platform (Windows, macOS, Linux)
  • Export to multiple CAD formats

Architecture:

C++ core with OpenGL rendering, Qt/QML for cross-platform UI, custom spline mathematics engine, real-time path optimization algorithms.

Full Tech Stack:

C++OpenGLQtQMLSpline Mathematics
Multiplayer Cycling Simulator Webapp

Multiplayer Cycling Simulator Webapp

Cross-platform cycling webapp with real-world Google Maps terrain and live users

Built a photorealistic cross-platform cycling simulator running on iOS, Android, PC, Mac, and tvOS. Generates accurate real-world maps on-the-fly using Google Maps integration, allowing users to ride real locations together in multiplayer. Features a performant Blur UI/UX design system consistent across all platforms, Bluetooth sensor support, and a social multiplayer layer for group rides.

UnityC#ASP.NET CoreSignalR+4 more
View Details

Key Features:

  • Real-world map generation from Google Maps API
  • Cross-platform: iOS, Android, PC, Mac, tvOS
  • Live multiplayer group rides
  • Bluetooth sensor integration (cadence, heart rate)
  • Performant Blur UI/UX design system across platforms
  • Location-based route discovery

Architecture:

Unity (C#) cross-platform client with Google Maps SDK terrain generation, ASP.NET Core backend for multiplayer session management and matchmaking, SignalR for real-time player state synchronisation, AWS for scalable hosting.

Full Tech Stack:

UnityC#ASP.NET CoreSignalRGoogle Maps APIAWSiOSAndroid
Oil Well Placement Training Simulation

Oil Well Placement Training Simulation

Gamified oil extraction training app with multi-variable scoring and geological hint overlays

Developed a Unity (C#) educational simulation that gamifies oil well placement training for industry trainees. Users place rigs based on geological hint layers — seismic lines, pressure gradient maps, and depth maps — and receive a calculated net earnings score driven by multi-variable equations factoring placement location, land value, and subsurface conditions.

UnityC#ASP.NET CorePostgreSQL+1 more
View Details

Key Features:

  • Multi-layer geological hint overlays (seismic, pressure, depth maps)
  • Multi-variable net earnings calculation engine
  • Gamified scoring with trainee performance tracking
  • Configurable land value and resource distribution maps
  • Designed for corporate training environments

Architecture:

Unity (C#) simulation and scoring engine, ASP.NET Core backend for trainee session management and score persistence, PostgreSQL for results and leaderboard storage.

Full Tech Stack:

UnityC#ASP.NET CorePostgreSQLWindows

Projects Impact

25+

Production Projects

5M+

Users Served

10K+

Concurrent Connections

99.5%+

Uptime Achieved