Mohd Waqar Azim

I'm an AI-focused Computer Science undergraduate with hands-on experience building production-grade ML and GenAI systems. My passion lies at the intersection of machine learning and software engineering—creating intelligent solutions that are not just technically impressive but genuinely useful.

Currently, I'm pursuing my B.Tech in Computer Science and Engineering (AI & Robotics) at Vellore Institute of Technology, Chennai. My work spans healthcare AI, computer vision, and RAG pipelines—from designing data pipelines and training models to building FastAPI backends and containerized deployments.

I've had the opportunity to work on real-world AI systems that process medical reports, analyze customer feedback at scale, and provide intelligence across codebases. My focus is always on building end-to-end systems that can handle real-world complexity and scale to production.

When I'm not coding, you'll find me exploring the latest developments in AI research, contributing to open-source projects, or working on my next ambitious side project.

1. Jun — Jul 2025

   Machine Learning Intern · Sugarbiz Consultant & Services Ltd London

   Built and trained CNN models to classify 50+ bird species from raw images, achieving 90%+ validation accuracy. Designed preprocessing and augmentation pipelines that improved generalization and reduced overfitting on noisy field data. Delivered an end-to-end ML pipeline (data prep → training → evaluation → inference) used in internal demos.

   • Python
   • TensorFlow
   • Deep Learning
   • CNN
   • Computer Vision

1. AdClass AI Platform

   AI-powered ad optimization platform trained on Google Ads, Meta, and TikTok data. Features 5 ML modules: Creative Performance Predictor, ROAS Optimizer, Client Churn Prediction, Multi-Touch Attribution Engine, and AI Audience Segmentation. Processes real-time ad data to turn insights into predictable revenue growth.

   • Python
   • Scikit-learn
   • XGBoost
   • FastAPI
   • Next.js
   • Docker
2. QuantXAlgo

   Production-ready algorithmic trading platform simulating a hedge fund's tech stack. Features a sandboxed Python IDE for quant strategies, backtesting engine with realistic execution, real-time portfolio dashboard, and risk management with VaR, stress testing, and kill switch functionality.

   • Python
   • NumPy
   • Pandas
   • FastAPI
   • Next.js
   • PostgreSQL
3. SentinelOS

   Production-grade AI platform for engineering and business intelligence across codebases. Implemented bug-risk and churn prediction using code metrics, TF-IDF, and ML models. Designed a knowledge graph and RAG-based LLM engine for CTO-style reports. Validated on real Google & Microsoft repositories.

   • Python
   • Scikit-learn
   • TF-IDF
   • NetworkX
   • OpenAI
   • Docker
4. Customer Intelligence AI

   Production-grade RAG system for analyzing large-scale customer feedback and generating business insights. Implemented TF-IDF + K-Means clustering and FAISS semantic search for intelligent information retrieval. Features LLM-powered insight generation with structured output validation.

   • Python
   • FAISS
   • Scikit-learn
   • LangChain
   • FastAPI
   • Docker
5. MedExplain

   Healthcare AI platform that parses multi-page medical reports in ~1-2s, extracting 40+ lab parameters. Designed deterministic explanation pipelines with safety filters, disclaimers, and confidence scoring. FastAPI backend handling 100+ concurrent requests with validation and rate limiting.

   • Python
   • PyTorch
   • OpenCV
   • FastAPI
   • Redis
   • Docker
6. Diagnostic Copilot

   Production-grade AI diagnostic tool using CheXNet for chest X-ray pathology detection with 95%+ accuracy. Built with torchxrayvision for authentic medical imaging analysis. Features robust error handling, image validation, and comprehensive medical disclaimers.

   • Python
   • PyTorch
   • CheXNet
   • torchxrayvision
   • FastAPI
   • Computer Vision
7. OpsAgent

   Enterprise-grade DevOps automation platform using AI for infrastructure analysis and zero-config deployment. ML-powered tech stack detection, cost optimization (50-70% savings), and blue-green/canary deployments with automatic rollback. 25+ database tables, 50+ API routes.

   • Go
   • Python
   • Next.js
   • PostgreSQL
   • Docker
   • Kubernetes
8. MediTranslate

   Healthcare translation platform enabling real-time doctor-patient communication. Features voice recording, automatic transcription via Whisper, translation in 20+ languages, and AI-powered medical summaries with severity detection and symptom extraction.

   • Next.js
   • TypeScript
   • Whisper
   • Groq API
   • Web Speech API
   • Vercel
9. NutriScan AI

   AI-powered calorie tracking app using computer vision. Snap a photo of food to get instant nutrition breakdown—no manual typing required. Features photo recognition, voice/text logging, and real-time nutritional analysis with macronutrient breakdown.

   • Python
   • Computer Vision
   • FastAPI
   • Next.js
   • TensorFlow
   • Docker
10. RedParrot

    Open-source AI interview copilot with real-time speech recognition via Whisper, AI answer generation using LLMs, resume intelligence parsing, and stealth mode for desktop. Features keyboard shortcuts and works with Groq API for fast inference.

    • Electron
    • React
    • TypeScript
    • Whisper
    • Groq API
    • LLMs
11. NEXUS AI Directory

    Production-ready AI tools directory featuring 50+ curated AI solutions across 10 categories. Advanced search and filtering with beautiful animations and world-class UI/UX. Fully responsive with performance optimizations.

    • Next.js
    • React
    • TypeScript
    • TailwindCSS
    • Framer Motion
    • Vercel
12. Bug Bounty Scanner

    Comprehensive security vulnerability scanner for web applications featuring automated reconnaissance, vulnerability detection, and detailed reporting. Implements industry-standard security testing methodologies.

    • Python
    • Requests
    • BeautifulSoup
    • Security
    • Automation