Amritsai

Crafting Solutions that matter.

Computer Science grad student at University of California, Irvine. When I'm not developing AI solutions & web applications, you'll find me geeking out over Tarantino's cinematography, debating JDM car legends, or analyzing football games. I specialize in AI, machine learning, and web development, focusing on building impactful solutions. Driven by curiosity and innovation, I'm always ready for the next technical challenge.

Background

Education, experience, and research that shaped my journey.

Experience

Bluebasket Labs

Software Development Intern

Coimbatore, IndiaDec 2024 – May 2025
15+vendors digitized via AI-powered SaaS platform
  • Migrated legacy .NET monolith to Microsoft IIS cloud architecture, boosting uptime by 30% and slashing memory by 50%
  • Architected full-stack platform with Next.js, Node.js, Supabase & Anthropic APIs for real-time customer engagement
  • Engineered CI/CD pipelines and monitoring with Docker, reducing deployment overhead by 20%
Education

University of California, Irvine

Master of Computer Science

Expected Dec 2026

AlgorithmsML & Data MiningAI
Education

Vellore Institute of Technology

B.Tech in Computer Science

Jun 2021 – Jul 2025

DatabasesCloud ComputingNetworks
Experience

Better Lives

AI Development Intern

Coimbatore, IndiaAug 2023 – Dec 2023
40%faster query resolution

RAG-based assistant with Python, RASA & OpenAI embeddings, serving 200+ daily warehouse operators

Published Research
View Paper

Performance Analysis of Deep CNN Models for Railway Track Damage Detection

Engineering Research Express (IOP Science)

95.42%accuracy · ResNet50

Comparative analysis of 7 CNN architectures for automated defect detection across 12,000 railway images from the Indian rail network

Featured Projects

A collection of projects that showcase my skills and experience.

SAP AI Supplier Sourcing Optimizer

Ongoing

Building an AI-driven supplier sourcing optimizer on SAP BTP that analyzes SAP S/4HANA purchasing history and supplier KPIs to recommend cost‑optimal vendor allocations across tariffs, logistics, and lead times. Includes interactive dashboards and what‑if tariff scenarios to compare historical vs. optimized spend.

SAP S/4HANA OData APISAP CAPSAP BTPSAP AI Core

Calorify

March 2026

AI-powered calorie and nutrition tracker for fitness. Uses image recognition (Google Gemini) to log meals from photos, an AI coach for personalized advice, and Supabase for auth and data. Built with Next.js, FastAPI, and deployed on Vercel and Railway.

Next.jsTypeScriptFastAPIPythonSupabaseGoogle Gemini

Hokansuru

Jan 2026

A full-stack cloud storage platform that utilizes Discord's infrastructure for unlimited free storage. Features end-to-end AES-256 encryption, intelligent file chunking, and a modern Discord-themed UI for seamless cross-device access and secure, privacy-preserving file sharing experiences.

GoNext.jsTypeScriptSQLite

Relnodes

Nov 2025

An interactive 3D network visualization platform that transforms LinkedIn connections into a stunning globe-based experience. Upload your professional network and explore geographic distribution, discover company-based referrals, and manage connections with intelligent tagging.

Next.jsTypeScriptReactPostgreSQL

PWagon

Apr 2024

A deep learning system for real-time license plate detection using a CNN trained on a custom dataset paired with Google's OCR. Adapted for Amber Alert networks and deployed on Raspberry Pi 4 for emergency monitoring.

TypeScriptJavaScriptPythonCSS

Pennywise

Feb 2024

An expense tracker that lets users log spending like a text message. Uses text categorization and NER to identify amounts, items, and automatically assign categories, then compiles insights to help users spot spending patterns.

JavaScriptKotlinGoDockerSpaCy

Tech Stack

The tools and technologies I use to bring ideas to life.

Get in Touch

Have a project in mind? Let's work together to create something amazing.