Connect on LinkedIn
Systems & AI Engineer

Dissecting technology beneath the surface

I'm Mohd Shahid. I build AI systems, construct robust Computer Vision pipelines, and design automated infrastructure layers. Driven by curiosity more than convention.

Interactive Playground Explore Projects

About Me

I work across AI, automation, computer vision, Linux, and infrastructure — building systems, debugging complexity, and understanding how things behave beneath the surface.

But my curiosity extends beyond machines.

I’m deeply interested in psychology, communication, language, and the invisible patterns that shape human behavior. I see technology and people as interconnected systems, both driven by structure, failure, adaptation, and intent.

Most of what I know was learned through experimentation: building, breaking, observing, and rebuilding. I value depth over trends, understanding over appearance, and questions over assumptions.

I’m not just interested in making systems work.

I’m interested in understanding why they work — technical systems, human systems, and everything in between.

shahid@core:~
$ system-diagnostics --verbose 
{
  "Name": "Mohd Shahid",
  "education": {
    "IIT Madras": "BS in Data Science (Expected Dec 2027)",
    "MANUU": "BTech in Computer Science",
    "AMU": "Diploma in Electrical Eng."
  },
  "certifications": [
    "Microsoft: AI-900, AZ-900, DP-900, PL-900",
    "NPTEL: PostgreSQL Cricket Analytics, IoT",
    "freeCodeCamp: APIs & Backend"
  ],
  "env": "Linux / Bash / PyTorch / Flask / DeepStream"
}
$

Featured Work

Dissecting models, compiling full-stack codebases, and automating administrative barriers.

Deep Learning & GenAI

Music Genre Classification

Developed a PyTorch deep learning pipeline to classify music genres. Extracted audio features (Mel-spectrograms, MFCCs) using Librosa and explored domain shift discrepancies by fine-tuning pretrained deep networks to handle noisy real-world composite signals, tracking experiments with Weights & Biases.

PyTorch Librosa Audio Processing W&B tracking Kaggle Competition
Modern Web App

Home Service v2

Built a full-stack, multi-role home services booking platform for the IIT Madras Modern Application Development 2 (MAD2) coursework. Gained hands-on experience implementing Celery for automated background tasks, Redis for caching, and SQLite for data management.

Flask Celery Redis SQLite3 Vue.js
AI vs Human Challenge

Paradox Text Classification

Competed in the IIT Madras Paradox competition to distinguish between human-written and AI-generated text. Engineered robust feature pipelines, tuned model hyper-parameters, and achieved 78.28% validation accuracy, ranking in the Top 7 participants.

Scikit-learn NLP Feature Engineering Text Classification
Fullstack Web

Flask Grocery Application II

Developed a secure web portal for buying and selling groceries. Integrated custom database designs, role-based authentication, and security configurations. Awarded an **'S' Grade** by IIT Madras.

Flask HTML/CSS Role-Based Auth S-Grade
Natural Language Processing

Financial Sentiment Analysis

Trained ComplementNB and MultinomialNB models on highly imbalanced financial comments. Optimized pipeline parameters to achieve a 70% test accuracy after thorough EDA.

Python ComplementNB EDA Imbalanced Data
Data Challenge

Customer Churn Prediction

Participated in the Coursera Data Science Challenge. Developed a classifier to predict customer churn in video streaming subscriber databases, achieving a 82.49% test accuracy and mapping user patterns to shape retention strategies.

Pandas NumPy Predictive Modeling Churn Analytics
Automation

Research Committee Management

Designed a database-driven system to automate administrative Research Advisory Committee (RAC) tasks at Maulana Azad National Urdu University. Successfully reduced manual logging delays.

Python SQLite3 Data Quality
Systems Core

rgb2hex Utility & Tools

Built modular libraries for fast color validations and coordinate tracking, utilizing Python OOP and comprehensive unit testing suites.

Python OOP Unit Testing BASH Utilities

Interactive ML Playground

Experience my AI models. Test my AI-vs-Human Text Classifier or try the Object Bounding Box simulator.

Paradox NLP Classifier

Enter a sentence or select a sample below to determine whether the writing is Human-written or AI-generated.

Classification Confidence
Human Written
96%
AI Generated
4%
Tokenizing input text stream...
Generating word distributions...
Running ComplementNB Classifier...
Output: Human Written (96% Confidence)
Awaiting text classification metrics...

Experience & Education

Sep 2025 - Present

Computer Vision Engineer

Nabeh | Full-time | Hyderabad, Telangana (On-site)

Contributing to real-world video analytics workflows. Hands-on development of edge computer vision pipelines using YOLO, NVIDIA DeepStream, and Kafka inside Linux environments to build scalable and maintainable features.

NVIDIA DeepStream YOLO Kafka Linux Video Analytics
Mar 2025 - Aug 2025

Computer Vision Intern

Nabeh | Internship | Hyderabad, Telangana (On-site)

Assisted in engineering computer vision pipelines, built edge streaming components under mentorship, and explored deep learning models for real-time video analysis.

Computer Vision Python Deep Learning Image Processing
Feb 2021 - Dec 2027 (Expected)

BS in Data Science and Applications

Indian Institute of Technology, Madras (IITM) | Roll No: 21F2000901

Intensive academic training in deep learning, data analytics, and machine learning structures. Completed the IITM Advanced Certificate in Machine Learning & Data Science. Earning Grade: A.

Deep Learning PostgreSQL Flask Pandas MLOps
Sep 2020 - Jul 2024

BTech in Computer Science

Maulana Azad National Urdu University (MANUU) | Hyderabad

Acquired rigorous foundations in computer systems, operating systems, data structures, algorithms, network protocols, and object-oriented programming. Developed database automation routines for University administration panels.

Computer Science Java Shell Scripting C SQLite3
2014 - 2017

Diploma in Engineering (Electrical)

Aligarh Muslim University (AMU) | Aligarh, UP

Acquired core principles of electrical engineering, hardware architectures, circuit calculations, signals, and control systems, framing my hardware-level understanding of computational structures.

Electrical Circuits Signals & Systems Hardware Foundations