Skip to main content
PROJECT ARCHIVE8 Documented Systems

Things I’ve Built

Every project was built to explore a real problem, test an engineering hypothesis, and learn from physical and digital failure modes.

01AIHardwareIoTEmbedded SystemscompletedFeatured

Kavach

AI-powered voice authentication and physical smart access system.

Core Builder & System Integrator — Voice biometric AI pipeline, Raspberry Pi server, Arduino serial protocol.
5-Person Team
The Problem

People frequently forget physical keys, while traditional contact biometrics require physical touch and specialized optical sensors. We asked: What if human voice could securely become the key?

The Solution & Idea

Turn human voice into a biometric key by integrating real-time deep learning speaker verification with a Raspberry Pi server, Arduino microcontroller, and electromechanical motor actuators.

Key Takeaway EarnedBuilding something real is fundamentally different from writing code that runs in a notebook.
ResemblyzerVoskWebRTC VADPythonFlask REST APIReactMongoDBJWT AuthRaspberry Pi 4Arduino UnoL298N Motor Driver
02SystemsBackendFinTechReal-TimecompletedFeatured

NanoTrade

Real-time paper trading platform with custom matching engine.

Solo Architect & Developer.
The Problem

Most trading demos are simple database dashboards that fail to simulate realistic order-book mechanics, execution latency, queue matching, and portfolio state reconciliation.

The Solution & Idea

Understand what happens underneath a real trading interface by building an event-driven paper trading platform powered by a custom matching engine.

Key Takeaway EarnedDeterministic event ordering and idempotent queues prevent race conditions in financial state machines.
C++ Matching EnginePythonFastAPIRedisCelerySupabase / PostgreSQLWebSocketsTypeScript
03AINLPRAGProductcompletedFeatured

SkillGap AI

Semantic embedding and RAG engine for resume and job description fit analysis.

Solo Builder & Product Architect.
The Problem

A résumé and a job description can superficially look like a match while still concealing critical capability gaps, leading to poor hiring signals and unfocused career preparation.

The Solution & Idea

Use semantic vector embeddings and Retrieval-Augmented Generation (RAG) to measure genuine competency fit, detect latent skill gaps, and generate structured learning roadmaps.

Key Takeaway EarnedAI is valuable when applied to domain-specific analytical problems rather than generic conversational prompts.
PythonFastAPIRAG & Vector EmbeddingsCeleryRedisReactSupabaseTypeScript
04MLOpsAICloudKubernetescompletedFeatured

MovieSentiment

Production-oriented ML discovery platform with automated MLOps & EKS deployment.

Solo Builder & MLOps Engineer.
The Problem

Machine learning models often remain trapped in Jupyter notebooks without robust CI/CD, containerized packaging, automated retraining, and scalable Kubernetes orchestration.

The Solution & Idea

Build an end-to-end ML sentiment discovery platform demonstrating the full lifecycle: Model → Version → Package → Deploy → Monitor.

Key Takeaway EarnedA machine learning model is only as dependable as the deployment pipeline and observability monitoring backing it.
PythonPyTorch / Scikit-learnMLflowDVCDockerKubernetesAmazon EKSGitHub ActionsPrometheusGrafana
05Machine LearningData AnalyticsCloudInternshipcompletedFeatured

Bike Demand Prediction ML System

Production ML demand forecasting application developed at TS Bridge.

Data Analytics & Machine Learning Intern (TS Bridge).
The Problem

Urban mobility fleets experience extreme utilization volatility due to weather, seasonal shifts, and commuting patterns, requiring precise predictive demand modeling to prevent asset shortages.

The Solution & Idea

Train, evaluate, and deploy an end-to-end regression pipeline on AWS with automated MLOps tracking and stakeholder analytics dashboards.

Key Takeaway EarnedModel accuracy metrics must translate directly into operational stakeholder decisions.
XGBoostScikit-learnPythonFlask REST APIDVCMLflowDockerAWS EC2AWS S3CI/CDPower BI
06Data ScienceAnalyticsInternshipcompleted

Large-Scale Data Preprocessing & EDA

Data wrangling and feature engineering pipeline on 50,000+ records at CourseVita.

Data Science Intern (CourseVita).
The Problem

Raw real-world datasets arrive laden with missing values, inconsistent encodings, extreme outliers, and duplicate records that degrade downstream modeling.

The Solution & Idea

Construct a robust, reproducible data wrangling and exploratory analysis pipeline to transform 50,000+ messy records into clean, modeling-ready datasets.

Key Takeaway EarnedReal-world data is inherently messy; systematic cleaning is the cornerstone of valid machine intelligence.
PythonPandasNumPyData PreprocessingMissing Value ImputationExploratory Data AnalysisMatplotlib / Seaborn
07Machine LearningMLOpsExperimentscompleted

Vehicle Insurance ML Pipeline

End-to-end predictive classification pipeline with modular MLOps architecture.

Solo Builder.
The Problem

Assessing vehicle insurance claim propensities requires handling severe class imbalances, multi-modal features, and repeatable retraining pipelines.

The Solution & Idea

Build a modular, reproducible ML pipeline implementing structured feature engineering, hyperparameter tuning, and containerized artifact tracking.

Key Takeaway EarnedModular code organization separates data engineering concerns from model evaluation.
PythonScikit-learnPandasFlaskDocker
08IoTSignal ProcessingExperimentscompleted

Industrial Sensor Fault Detection

Anomaly detection system for industrial sensor telemetries.

Solo Builder.
The Problem

Industrial machinery sensor drifts and electrical spikes often go undetected until catastrophic mechanical failure occurs.

The Solution & Idea

Apply time-series statistical anomaly detection and threshold heuristics to continuous sensor telemetry streams.

Key Takeaway EarnedDomain knowledge of electrical signals is critical for designing effective anomaly filter thresholds.
PythonNumPyTime-Series Signal AnalysisIoT Protocols