Shaik Mohammad Abrar

About

You should add a summary here

Work Experience

Zelestra

Invalid Date NaN - Present

Hackathon Zelestra

Engineered and optimized a regression pipeline, leveraging ensemble models (XGBoost, LightGBM, Random Forest) to predict solar panel efficiency from real-world energy data. Applied feature engineering and hyperparameter tuning techniques, significantly reducing RMSE and achieving 89.87% prediction accuracy.

Amazon

Invalid Date NaN - Present

Amazon ML Challenge

Developed a multi-label image tagging system for e-commerce, utilizing Paddle OCR and deep learning to precisely extract and analyze visual/textual product content. Achieved a high F1-score of 0.92, enabling accurate product categorization and improving catalog visibility and search relevance for sellers.

Education

Abdur Rahman Crescent Institute of Science and Technology

—

Bachelor of Technology - Artificial Intelligence and Data Science

Sri Chaitanya Junior College

—

12th Standard

Skills

Python
JavaScript
React
Node.js
Machine Learning
NLP
AWS
MongoDB
SQL
Data Analysis

My Projects

Check out my latest work

I've worked on a variety of projects, from simple ones to complex. Here are a few of my favorites.

AI-Powered Journaling & Habit Tracking Web App

Invalid Date NaN - Present

Built a full-stack application using React, MongoDB, and AWS, allowing users to log daily entries and monitor personal habits. Implemented an AI module (e.g., using NLP for sentiment analysis/topic modeling) to analyze journal data, providing personalized productivity suggestions and improving user consistency and self-improvement habits.

Real-time Competitive Programming Battle Platform

Invalid Date NaN - Present

Designed and developed a scalable, real-time coding battle platform (inspired by LeetCode), enabling matched users to compete on CP problems with dynamic, chess-style rating updates. Engineered with React, Express.js, and AWS services for real-time matchmaking, secure user management, and scalable hosting.

AI-Enhanced Video Sharing Platform (YouTube-inspired)

Invalid Date NaN - Present

Developed a comprehensive video-sharing platform using Next.js and Mux, integrating Gemini for automated transcription and SEO metadata generation. Engineered core features including playlists, comments, analytics, and authentication, with a responsive UI built using Tailwind CSS, Shadcn UI, and Drizzle ORM.

AI Voice Synthesis Application (ElevenLabs-inspired)

Invalid Date NaN - Present

Constructed a full-stack AI voice synthesis application using Flask and React, leveraging the ElevenLabs API for text-to-speech conversion. Implemented features such as custom voice cloning, secure file uploads, multi-voice playback, and a responsive user interface with authentication.

Contact

Get in Touch

Write some text here... maybe a catchy phrase for people to contact you? 👀