Maaz Ahmed
Python Automation | AI Solutions | Data Analytics
New to Contra
Maaz is building their profile!
AgriVision: AI-Powered Agricultural Platform
I developed AgriVision, a modern web application designed to empower farmers and agricultural businesses. This platform integrates real-time crop monitoring and seamless trade workflows into a secure, user-friendly ecosystem.
At its core, AgriVision features an integrated AI-driven disease detection tool (powered by a custom YOLOv8} that instantly analyses crop health from uploaded images. By combining a clean, intuitive frontend interface with advanced backend machine learning, AgriVision helps users secure better yields and manage their workflows efficiently.
Key Features & Tech Stack:
Full-Stack Web Development with secure user authentication
Integrated Computer Vision for real-time disease detection
Python backend and java frontend
1
17
Interactive Global Retail Sales Dashboard
I designed and developed this interactive data visualization dashboard to transform raw global retail data into a clear, actionable business intelligence tool.
Instead of relying on static spreadsheets, executive stakeholders can use this dashboard to instantly view high level KPIs. This centralized visualization helps management identify top-performing sectors, track seasonal trends, and drive data informed strategic decisions.
Key Features & Tech Stack:
Tableau for advanced data visualization and interactive dashboard design
Geospatial Analysis for mapping global sales distributions
Dynamic Filtering for custom date ranges, regions, and product categories
Business Intelligence & Data Analytics for accurate KPI tracking
1
20
Custom YOLOv8 Plant Disease Detection Model
I trained and deployed a custom computer vision model designed to automate agricultural crop inspections and monitor plant health in real-time.
Using a custom-trained YOLOv8 algorithm, this system accurately identifies and classifies 12 distinct plant diseases (including Bean Rust, Gray Mold, and Spider Mites) from raw field images. The model generates precise bounding boxes and confidence scores, providing an automated solution that drastically reduces manual inspection time and helps improve overall crop yields.
Tech Stack & Architecture:
YOLOv8 for high-speed, real-time object detection
Python backend for data processing and inference scripting
Deep Learning & Neural Networks trained on custom agricultural datasets
Model Evaluation & Optimization (mAP tracking and loss analysis)
1
24
Custom RAG AI Research Assistant
This is a custom Retrieval-Augmented Generation (RAG) AI tool I developed to instantly parse and answer complex questions from dense academic papers and corporate documents.
Instead of manually reading through massive PDFs, users can query this AI to instantly extract accurate, context-aware answers. To ensure zero hallucinations and build trust, the AI strictly sources its responses from the provided literature and actively cites its references.
Tech Stack & Architecture:
LangChain for agent orchestration and data retrieval
Hugging Face LLMs for intelligent inference
Vector Databases for fast semantic search
Python backend integrated seamlessly with a clean, user-friendly frontend
1
30