Skip to content
Open to work

AI & ML Engineer

Building intelligent systems that · solve real-world problems.

Explore Work
01 / Calibration

I Build AI Systems That Solve Real-World Problems.

I'm an AI Engineer passionate about Machine Learning, Deep Learning, Computer Vision, NLP, and Generative AI. I focus on building production-ready AI solutions that combine research with real-world impact.

CurrentlyBS AI Student — Riphah International University
02/Selected work

Four systems, built beyond the classroom.

Each project below started with a real problem, shipped as a working system, and left a measurable trace.

Case studies
04
2024 — 2026
01
Case / 2026 · Hackathon

CropHeat AI

Hyperlocal climate intelligence turned into explainable, crop-specific heat-risk decisions.

CropHeat AI — cover
● Case / 01
Solo build
Hackathon
CropHeat AI
Risk score
0–100
Risk factors
6
Hackathon
FortyGuard
The problem

Farmers and agronomists get generic weather forecasts, but a headline temperature does not explain whether heat is dangerous for a specific crop at a specific growth stage.

The solution

Built a FortyGuard-powered platform that combines spatial temperature, exceedance and persistence data with sourced crop heat-sensitivity thresholds to produce a transparent 0–100 risk score and actionable advisory.

Technology
Next.jsFastAPIFortyGuardLeafletSQLite
Read case study View source
Next case
02
Case / 2026 · Shipped

PlantScan

Three CNN backbones, compared and shipped, for plant disease diagnosis.

PlantScan — cover
● Case / 02
Semester Project
Shipped
PlantScan
Accuracy
98%
Classes
38
Models compared
3
The problem

Smallholder farmers lose crops to diseases they can't identify fast enough — lab tests are slow, expensive, and rarely reachable.

The solution

Trained and compared EfficientNetB0, MobileNetV2 and ResNet50 via transfer learning, then shipped the comparable models behind a Flask app with live camera capture and quality-gated auto-capture.

Technology
TensorFlowKerasEfficientNetB0FlaskOpenCV
Read case study View source
Next case
03
Case / 2026 · Shipped

Smart Electricity Consumption

Predicts household electricity use, with real energy-saving recommendations.

Smart Electricity Consumption — cover
● Case / 03
Teyzix Core Internship Project
Shipped
Smart Electricity Consumption
R² score
0.964
MAE
1.57 kWh
Models compared
7
The problem

Households can't see where their energy goes until the bill lands — too late to change anything.

The solution

Compared 7 regression models on a household dataset, then shipped the best-performing model behind a Flask app that estimates consumption and gives savings recommendations.

Technology
PythonScikit-learnXGBoostPandasFlask
Read case study View source
Next case
04
Case / 2026 · Academic

EduPredict

A full-stack university portal built around structured student, course, result and fee workflows.

EduPredict — cover
● Case / 04
Semester Project
Academic
EduPredict
Portals
2
Database
MySQL
Backend
Flask
The problem

University records span students, courses, enrollments, marks and fees; without a structured system, these workflows become fragmented and difficult to manage.

The solution

Built a Flask + MySQL portal with separate admin and student workflows, session-based authentication, course enrollment, automatic grade calculation, published results, CGPA and semester-fee management.

Technology
PythonFlaskMySQLHTML/CSSJavaScript
Read case study View source
Other builds

Smaller systems, still shipped.

Earlier experiments stay visible without competing with the four projects that best represent my current engineering range.

Jarvis AI Assistant preview
2025 · Other build

Jarvis AI Assistant

A multilingual voice assistant powered by Gemini, with real-time Hindi-to-English translation.

GitHub
RoboImmi preview
2025 · Other build

RoboImmi

A desktop AI chatbot for DSA and AI questions, by voice or text.

GitHub
02/About

I design and ship production-grade AI systems — from data pipelines to inference. My focus is computer vision, generative agents, and applied ML that survives contact with real users. I care about latency, evaluation, and interfaces that make intelligence feel effortless. Currently focused on my studies and open to internship opportunities in AI and machine learning.

Based in
Lahore, Pakistan
What I Offer
Machine Learning Engineering
AI Engineering
Data Science
Status
Open to Internships
03/Trajectory Engineering log

An engineering path, logged like commits.

Every internship is a checkpoint — real teams, real datasets, systems shipped end-to-end.

Currently
Pursuing Bachelor's in Artificial Intelligence
GCI World 2026 Participant — Matsuo & Iwasawa Laboratory, The University of Tokyo
Roles
03
Log #01ML capstone: stress prediction
16 July 2026 — 28 August 2026

Machine Learning Intern

NeuroFive Solutions

Completed the ML Track capstone: an end-to-end stress-level prediction pipeline — data cleaning, feature engineering, and a three-model comparison (Linear Regression, Random Forest, XGBoost) — deployed as a live Streamlit app, with an honest write-up of the dataset's limitations.

Models compared
3
Dataset
374 records
Status
Completed
PythonScikit-learnXGBoostPandasStreamlit
View repository
Log #02Energy forecasting system
11 June 2026 — 10 July 2026

Machine Learning Intern

Teyzix Core

Developed an end-to-end Smart Electricity Consumption Prediction system using regression models, feature engineering, data analysis, and Flask deployment to provide intelligent energy optimization recommendations.

R² score
0.964
MAE
1.57 kWh
Shipped
1 system
PythonScikit-learnPandasFlaskMachine Learning
Log #03AI workflows & automation
10 June 2026 — 10 July 2026

AI Intern

Decode Labs

Worked on AI applications, machine learning workflows, and intelligent automation while gaining practical experience in solving real-world AI problems.

Domain
Automation
Stack
Python · ML
Duration
1 month
PythonArtificial IntelligenceMachine LearningAutomationGit
04/Technical Architecture

Behind the intelligence.

A live look at the systems behind the systems — repositories, stack, certifications, and internships wired together like a dashboard.

01 · GitHub dashboard
Imran Ghafoor · GitHub avatar
@Imran-Ghafoor594
Imran Ghafoor

AI Engineer building ML, deep learning, and AI-powered systems.

Repositories
25
public
Stars earned
19
across repos
Focus
AI / ML
primary
Languages
Aggregate
Python65%
Jupyter Notebook27%
Others8%
Visit profile Browse repos
02 · Tech stack
Languages
03 items
PythonSQLMongoDB
AI & Machine Learning
06 items
TensorFlowPyTorchScikit-learnOpenCVHugging FaceComputer Vision
Frameworks
02 items
FlaskFastAPI
Tools
05 items
GitGitHubVS CodeGoogle ColabJupyter Notebook
03 · Certifications
Verify credentials
06/Transmission

Let’s build something intelligent together.

I’m always interested in AI, Machine Learning, Generative AI, and Computer Vision opportunities where technology creates meaningful impact.

GitHub
Imran-Ghafoor594
Visit GitHub✓ Copied
LinkedIn
in/imranghafoor56
Open LinkedIn✓ Copied
Resume
Imran_Ghafoor.pdf
Download Resume✓ Copied