Complete project catalog with major builds, hackathon work, and learning projects.
Click a project title to open its video preview
Scalable MLOps-driven market prediction system with real-time analytics, automated retraining, and cloud-native deployment.
Engineered a production-grade financial forecasting platform processing 1M+ market data points with automated retraining pipelines, drift-aware evaluation, and low-latency cloud deployment.
MLOpsTime-Series ForecastingCI/CDMicroservicesCloud InfrastructureDistributed Systems
Built cloud-native forecasting platform with automated training, evaluation, and deployment workflows
Designed modular microservice architecture for scalable AI inference and analytics pipelines
Implemented CI/CD pipelines with containerized deployment and production-grade infrastructure setup
Integrated drift monitoring and retraining workflows for adaptive financial prediction systems
Optimized real-time inference pipeline for low-latency forecasting and scalable API serving
Ask this project
Scoped to ForesightX - The Future Has Signals. We Decode Them.
AI-powered ECG interpretation, remote consultations, and real-time healthcare intelligence in a unified clinical platform.
Developed a full-stack healthcare AI ecosystem achieving 91.3% F1 on ECG benchmarks with sub-150ms inference alongside integrated telemedicine, patient workflows, and diagnostic assistance.
Healthcare AIDeep LearningSignal ProcessingReal-time SystemsFull Stack Development
Built AI-driven healthcare platform combining ECG diagnostics, telemedicine, and patient monitoring
Integrated transformer-based ECG classification pipeline with near real-time clinical inference
Implemented secure video consultations, digital health records, and AI-assisted diagnostic workflows
Designed scalable backend services and healthcare-focused system architecture for responsive care delivery
Developed anomaly detection and rhythm classification pipelines using custom ECG preprocessing workflows
Ask this project
Scoped to PulseCare AI - Intelligent Cardiac Diagnostics & Telehealth Platform
LLM-powered interview automation with adaptive questioning, candidate evaluation, and recruiter-ready analytics.
Engineered an AI recruitment ecosystem capable of generating resume-aware interviews, evaluating candidate responses, and producing structured hiring insights through automated workflows.
Developed AI-powered interview engine with resume-aware and job-description-aware adaptive questioning
Built end-to-end recruitment workflow covering candidate onboarding, interviews, evaluation, and reporting
Implemented automated recruiter scoring and structured performance analytics using LLM-based evaluation
Integrated speech processing and conversational AI workflows for interactive interview experiences
Designed scalable backend APIs and database workflows for recruiter and candidate management
Ask this project
Scoped to IntelliHire AI - Autonomous Recruitment & Interview Intelligence Platform
AI-driven resume tailoring, ATS optimization, and smart job matching for modern hiring workflows.
Built an end-to-end career intelligence platform enabling automated resume customization, ATS analysis, and AI-assisted job discovery with production-ready full-stack deployment.
Generative AILLMsATS OptimizationPrompt EngineeringFull Stack Development
Developed AI-powered resume customization based on job descriptions using LLM-driven workflows
Implemented ATS scoring and keyword optimization to improve recruiter-facing resume quality
Built intelligent job-role matching system with semantic resume analysis and recommendation logic
Designed responsive full-stack platform with real-time document processing and dynamic resume generation
Integrated automated career assistance workflows for resume refinement and application preparation
Ask this project
Scoped to CareerPilot AI - Intelligent Resume & Job Matching Platform
Multimodal computer vision pipeline for automated drywall crack and seam segmentation using prompt-aware deep learning.
Engineered a reproducible end-to-end industrial AI pipeline achieving 0.3643 macro IoU and 0.5055 Dice Score using CLIPSeg + CNN refiner architecture with prompt-conditioned segmentation and experiment-tracked ML workflows.
Computer VisionSemantic SegmentationMultimodal AIMLOpsIndustrial AI
Built prompt-conditioned segmentation pipeline for drywall crack and seam detection using CLIPSeg and SAM
Designed reproducible ML workflow with DVC, MLflow, and DagsHub for experiment tracking and evaluation
Implemented pseudo-mask generation pipeline for weakly supervised segmentation from box-only datasets
Optimized CLIPSeg + CNN refiner architecture with Focal Tversky loss for thin-structure defect segmentation
Developed automated binary mask export system with benchmark evaluation using IoU and Dice metrics
Ask this project
Scoped to VisionInspect - Prompt-Conditioned Industrial Defect Segmentation System
Drone-integrated waste detection with YOLO.
Built a drone-assisted detection and routing workflow that streamlines waste monitoring and cleanup operations.
Computer VisionAutomationSystem Design
Drone image ingestion pipeline for field conditions
YOLO-based object detection integrated with backend workflows
Geo-mapped outputs to support operational decisions
Ask this project
Scoped to BinSavvy: Smart Waste Management System
Coach tracking, food discovery, and trip planning.
Developed a full-featured travel assistant web app focused on real-time utility and reliable fallback behavior.
Full-StackProduct EngineeringUX
Built coach tracking and trip planning flows end-to-end
Integrated public schedule and station datasets
Shipped user-focused interfaces for quick mobile usage
Ask this project
Scoped to MargDarshak-Mitr: Railway Travel Assistant
Predict GPA and stress levels using lifestyle signals.
Built a 2,000-record data science pipeline with feature engineering, cross-validation, and regression modeling.
Data ScienceRegressionModel Evaluation
Structured preprocessing and feature engineering workflow
Compared Linear Regression, Random Forest, and SVR
Produced interpretable insights for stakeholder communication
Ask this project
Scoped to Student Lifestyle and Academic Performance Predictor
Cuisine popularity prediction over 10,000+ records.
Implemented an end-to-end analytics and modeling pipeline with ~92% classification accuracy.
Data AnalyticsClassificationVisualization
Processed and analyzed 10,000+ global restaurant records
Trained Random Forest and XGBoost models for cuisine popularity
Built trend visualizations for city-wise and rating-based insights
Ask this project
Scoped to Restaurant Data Analytics & Cuisine Popularity Prediction
Chess Game - Two-Player Board Game
Interactive two-player chess with core mechanics.
Strengthened software engineering fundamentals through object-oriented game-state modeling.
Software EngineeringOOPProblem Solving
Implemented piece movement and turn-based game flow
Modeled board state and rule validation logic
Designed clean class-based architecture for maintainability
Ask this project
Scoped to Chess Game - Two-Player Board Game
Contact Me
Let's connect
Have a project in mind or want to discuss opportunities? I’d love to hear from you.
I’m open to software engineering, data science, and ML engineering roles — and new collaborations.