Try the Interactive Camera

Enable your camera to experience the live pixel-grid effect.

Video is processed locally and never leaves your device

AI & Data Engineer

Building intelligent systems, autonomous data pipelines, and scalable backend architectures. Transforming complex data into actionable intelligence.

Manus-Macbook — bash

Featured

Vinaval AI

AI-Powered Competitive Exam Preparation Platform

Completed

An AI-powered learning platform for competitive exam preparation combining curriculum-grounded tutoring, document-based study assistance, streaming quiz generation, and progressive flashcards in English and Tamil.

Notable Features

•AI Tutor
•My Study GPT
•TN Textbook RAG

ReactVitePythonFastAPI+13 more

View Source Code

Details →

StreamFlow

Real-Time Event Streaming & Distributed Data Engineering Pipeline

Completed

A local real-time e-commerce data engineering pipeline streaming synthetic order events through Kafka, validating and processing them with PySpark Structured Streaming, and writing to PostgreSQL with dead-letter queue isolation.

Notable Features

•Synthetic Event Generation
•Kafka Decoupled Ingestion
•PySpark Structured Streaming

Python 3.12Apache Kafka 3.7PySpark / Apache Spark 3.5 Structured StreamingPostgreSQL 16+5 more

View Source Code

Details →

LedgerBud

Personal Finance Intelligence & Wealth Management Platform

Completed

A personal finance intelligence platform for managing wallets, transactions, budgets, savings goals, and recurring payments with automated financial health scoring, analytics, and an AI financial advisor.

Notable Features

•Multi-Wallet Management
•Automated Statement Parsing
•Financial Health Scoring

StreamlitPlotlyPythonFastAPI+13 more

View Source Code

Details →

Churn Prediction

Telecom Customer Retention Intelligence & Survival Analysis

Completed

An ML-powered telecom customer retention dashboard combining CatBoost classification, Cox proportional-hazards survival analysis, composite risk scoring, and automated retention recommendations with batch CSV prediction.

Notable Features

•CatBoost Classifier
•Logistic Regression Baseline
•Cox Proportional-Hazards Analysis

PythonStreamlitCatBoostscikit-learn+6 more

Live Demo View Source Code

Details →

FieldMind

AI-Powered Multi-Stage Agricultural Assistance & Computer Vision Platform

Completed

An educational agricultural assistance platform combining multi-stage computer vision inference for leaf disease detection, soil-based crop recommendations, diagnosis history, and a farmer equipment marketplace.

Notable Features

•Stage 1 — Leaf Verification
•Stage 2 — Disease Classification
•Stage 3 — Severity Detection

React 19ViteTailwind CSSFramer Motion+20 more

View Source Code

Details →

Broklink

Full-Stack Rental Property Discovery & Management Platform

Completed

A full-stack rental property platform enabling users to register profiles, publish property listings, update rental availability, and explore listings through dynamic location-based cascading search and filtering.

Notable Features

•Session Authentication & Profiles
•Listing Management Lifecycle
•Cascading Location Filters

PythonFlask 3.1.2SQLAlchemy 2.0.43Flask-SQLAlchemy+8 more

View Source Code

Details →

Engineering

Building intelligent applications, scalable backend systems, and robust data engineering architectures.

Programming Languages

Python

SQL

Java

JavaScript

TypeScript

Data Engineering & Distributed Processing

Apache Kafka

PySpark

PostgreSQL

Apache Airflow

ETL & Data Quality

Docker Compose

Backend & API Engineering

FastAPI

Flask

SQLAlchemy

Pydantic

Authentication

AI, ML & Computer Vision

PyTorch

ONNX Runtime

YOLOv8

CatBoost

Cox PH Survival

Scikit-learn

OpenCV

Generative AI & LLM Integration

Google Gemini

Groq API

RAG & Vector Search

Server-Sent Events (SSE)

Document Ingestion

Databases & Persistence

SQLite

PostgreSQL

MySQL / MariaDB

ChromaDB

Google Firestore

Frontend & Analytics

React

Vite

Streamlit

Plotly

Power BI

Tailwind CSS

Testing & DevOps

Pytest

Docker

Git & GitHub

Selenium

Playwright

Applied

Vinaval AI

FastAPI, RAG, ChromaDB, Gemini/Groq, SSE Streaming, React

LedgerBud

FastAPI, Streamlit, PostgreSQL 15, Double-Entry, Plotly, Groq AI

Churn Prediction

CatBoost, Cox PH Survival, Composite Risk, Streamlit, Groq

Broklink

Flask 3.1.2, SQLAlchemy, MySQL, Session Auth, Cascading Filters

StreamFlow

Apache Kafka, PySpark Structured Streaming, PostgreSQL, Docker

FieldMind

Multi-Stage CV (ONNX, YOLOv8, EfficientNet), FastAPI, Firestore, React

Global Impact &

Designing systems that scale across borders. From high-throughput streaming pipelines to robust machine learning models, my engineering focus is on reliability, efficiency, and worldwide accessibility.

Resilient Data Pipelines (Kafka, Spark)

Scalable Microservices (FastAPI, Docker)

Performant AI Integration (RAG, Vector DBs)

Journey &

A combination of competition success, consistent learning, and practical engineering.

KALAM 2025 — Meme Pudi

Sri Shakthi Institute of Engineering and Technology

🥇 1st Place

Won first place in Meme Pudi, a creative team event involving meme-dialogue recognition and expressive acting. The event brought together 28 teams and tested creativity, quick thinking, and team coordination.

Team Collaboration:

Sachin Annadasan A G

View LinkedIn Post

KALAM 2025 — Algo Arena

Sri Shakthi Institute of Engineering and Technology

🥈 2nd Prize

Secured second place in a three-round coding challenge organized by the Department of Information Technology.

The competition included:

Round 1: C and Data Structures multiple-choice questions.
Round 2: Debugging challenge.
Round 3: Live coding round.

The event tested programming fundamentals, debugging skills, and coding speed.

View LinkedIn Post

ALTRUIX 2025 — Data Fest

Nehru Institute

🥉 3rd Prize

Secured third place in Data Fest through hands-on work with a practical data science workflow involving data preprocessing, data handling, visualization, model training, and accuracy evaluation.

The experience strengthened collaborative problem-solving and applied data science skills.

Team Collaboration:

Monika M Haripriya Senthil Kumar

View LinkedIn Post

Smart India Hackathon 2025

Semi-Finalist

FieldMind demonstrates the integration of deep learning, computer vision, and full-stack web development in an agricultural assistance platform. The application combines AI-driven crop disease analysis, crop recommendations, and a peer-to-peer agricultural marketplace within a responsive web application.

Key capabilities

1. AI-Powered Crop Disease Detection:Implemented a multi-stage inference pipeline that combines leaf verification to filter obvious non-leaf inputs, disease classification to identify possible disease categories, and YOLO-based severity detection to locate visible symptom regions.
2. Intelligent Crop Recommendations:Uses soil and environmental parameters such as nitrogen, phosphorus, potassium, pH, temperature, humidity, and rainfall to generate crop recommendations.
3. Agricultural Marketplace:Provides a platform for equipment listings and agricultural workforce services, including booking workflows.
4. Full-Stack Application:Built a React frontend and FastAPI backend, integrating Firebase Authentication and Firestore for authentication and data persistence.

Project significance

FieldMind demonstrates how separate machine-learning components can be orchestrated behind a REST API and integrated into a complete application with authentication, persistent data, and a responsive user interface.

Important transparency

The FieldMind README explicitly describes the project as an educational and demonstration project rather than a production-ready or agronomically validated system.

Technology stack

PythonFastAPIReactPyTorchONNX RuntimeYOLOv8scikit-learnFirebase AuthenticationGoogle Firestore

View source code

Team Collaboration:

Sanjay Nithin S. Shanmathy Narayanan Subhashree R. Arun S. Hari Prasath

Lean Six Sigma Yellow Belt

96% Assessment Score

Earned the Lean Six Sigma Yellow Belt certification with a score of 96%, building a foundation in process improvement, quality management, and operational efficiency. The training covered Lean and Six Sigma principles, practical applications, defect analysis, and quality-improvement methodologies. A case study involving defect calculations in Cadbury Gems helped connect the concepts to a real-world-style quality assessment.

Key takeaways

Efficiency and defect reduction:Understanding how Lean and Six Sigma support efficiency and defect reduction.
Structured problem-solving:Applying structured problem-solving and process-analysis techniques.
Continuous improvement:Recognizing opportunities for continuous improvement and cost optimization.
Operational decision-making:Building business awareness and a stronger foundation in operational decision-making.

View LinkedIn Post

Certifications &

500+ Problems Solved

LeetCode

Algorithmic Problem Solving

Top 11.97% Globally

LeetCode

Competitive Programming

SQL for Data Science

Coursera / IBM

SQL Basic

HackerRank

SQL Intermediate

HackerRank

Python Basic

HackerRank

Infosys Springboard Virtual Internship 7.0

September–November 2026

Ongoing

Selected for the Infosys Springboard Virtual Internship 7.0, working on a full-stack platform that helps organizations create, localize, personalize, distribute, and monitor public-awareness campaigns.

Key responsibilities and features

AI-assisted content generation:Developing tools to generate campaign content using large language models.
Multilingual translation:Supporting translation and localization for Indian-language audiences.
Audience segmentation:Organizing audiences for targeted communication.
Campaign management:Building workflows for campaign creation, scheduling, and personalization.
Multi-channel communication:Supporting campaign workflows for email, SMS, WhatsApp, push notifications, and web broadcasts, as applicable to the implementation.
Engagement analytics:Tracking campaign engagement and presenting useful insights.
Sentiment and compliance:Incorporating content analysis to support appropriate tone and compliance checks.

Technology stack

FastAPIReact.jsLarge Language Models (LLMs)

View LinkedIn post View source code

Intel Unnati Industrial Training

Intel Unnati

October 06, 2025 – December 20, 2025

Designed and developed VisionRec, a visual search and recommendation system that identifies similar fashion items using image embeddings, vector similarity search, and knowledge-graph-based recommendations. VisionRec moves beyond traditional keyword-based search by analyzing the visual meaning of images. It supports image and text-based similarity search, recommendations, and semantic relationships between fashion products.

Key capabilities

Image and text similarity search:Used CLIP to represent images and text in a shared embedding space.
Optimized inference:Integrated Intel OpenVINO for efficient model inference.
Vector similarity search:Used FAISS to retrieve visually similar items.
Knowledge-graph recommendations:Connected related products through semantic relationships.
Out-of-distribution detection:Added a mechanism to reject non-fashion objects.
Authentication and data management:Used JWT authentication and Supabase PostgreSQL.
Interactive graph visualization:Used D3.js to display product relationships.
Backend and frontend:Implemented a Flask backend with a lightweight JavaScript frontend.

Dataset

The system used a curated fashion dataset containing 32,356 images across 29 fashion categories.

Key learning

The project reinforced that an intelligent AI application involves more than training a model. Inference optimization, vector retrieval, data management, recommendation logic, security, visualization, and user experience must work together.

Technology stack

PythonFlaskCLIPIntel OpenVINOFAISSSupabase PostgreSQLJWTD3.jsJavaScript

Team Collaboration:

Sanjay Nithin S Krishnakumar T

View LinkedIn post View source code

Cloud Developer Intern

Career Ladder

September 2025 – October 2025

Worked with AWS EC2, RDS and S3 to deploy multi-tier cloud applications.
Used Docker to containerize applications and maintain consistent development and deployment environments.
Gained hands-on experience with cloud infrastructure, backend deployment and cloud resource management.

B.Tech — Artificial Intelligence & Data Science

Sri Shakthi Institute of Engineering and Technology (SIET), Coimbatore

2023 – 2027

Current CGPA: 8.44 Status: Pursuing; expected graduation in 2027. Pursuing Artificial Intelligence and Data Science, with hands-on experience building AI-powered applications, backend systems, and data engineering pipelines. Developed projects spanning generative AI, machine learning, computer vision, and real-time data processing, with a focus on translating technical concepts into practical solutions.

HSC / Class XII

Government Higher Secondary School

2023

Score: 86% Achieved 86% while developing leadership and communication skills through academic and collaborative activities. Built a strong interest in Computer Science and Chemistry, motivating me to explore the intersection of technology, logical problem-solving, and scientific thinking.