You will get Custom AI-Powered Applications for Your Business
Project details
I build production-grade AI & Deep Learning systems from
research prototypes to fully deployed, containerized solutions.
My specializations:
✅ Computer Vision & Object Detection (YOLO, OpenCV)
✅ Medical Imaging AI (histopathology, cancer classification)
✅ Custom CNN & Transfer Learning (ResNet, EfficientNet, DenseNet)
✅ FastAPI + Docker AI backends & REST APIs
✅ On-device / Edge AI (TensorFlow Lite)
✅ Model Explainability (Grad-CAM, SHAP, saliency maps)
What makes me different:
I don't just hand you a notebook I deliver clean, documented,
deployable code with real evaluation metrics (Accuracy, F1,
AUC-ROC, Confusion Matrix) and Grad-CAM visualisations so you
actually understand what the model is doing.
Recent projects I've shipped:
→ Blood cancer multi-class classifier (ResNet-50, EfficientNet-B3)
→ Real-time traffic violation detection system (YOLOv8 + FastAPI)
→ On-device fruit AI scanner (TensorFlow Lite — Google Play Store)
→ Full-stack Django web app (Docker, live on Hugging Face)
PEC Registered Engineer | B.Sc. Software Engineering | 2+ years
teaching AI to 100+ students.
Let's build something that actually works.
research prototypes to fully deployed, containerized solutions.
My specializations:
✅ Computer Vision & Object Detection (YOLO, OpenCV)
✅ Medical Imaging AI (histopathology, cancer classification)
✅ Custom CNN & Transfer Learning (ResNet, EfficientNet, DenseNet)
✅ FastAPI + Docker AI backends & REST APIs
✅ On-device / Edge AI (TensorFlow Lite)
✅ Model Explainability (Grad-CAM, SHAP, saliency maps)
What makes me different:
I don't just hand you a notebook I deliver clean, documented,
deployable code with real evaluation metrics (Accuracy, F1,
AUC-ROC, Confusion Matrix) and Grad-CAM visualisations so you
actually understand what the model is doing.
Recent projects I've shipped:
→ Blood cancer multi-class classifier (ResNet-50, EfficientNet-B3)
→ Real-time traffic violation detection system (YOLOv8 + FastAPI)
→ On-device fruit AI scanner (TensorFlow Lite — Google Play Store)
→ Full-stack Django web app (Docker, live on Hugging Face)
PEC Registered Engineer | B.Sc. Software Engineering | 2+ years
teaching AI to 100+ students.
Let's build something that actually works.
AI Development Type
Deep Learning, Knowledge Representation, Model Tuning, Recommendation System, Software MaintenanceAI Tools
Keras, MLflow, NVIDIA AI Platform, Open Neural Network Exchange, OpenCV, PyTorch, TensorFlowAI Development Language
PythonWhat's included
| Service Tiers |
Starter
$100
|
Standard
$220
|
Advanced
$525
|
|---|---|---|---|
| Delivery Time | 7 days | 15 days | 24 days |
Number of Revisions | 1 | 3 | 5 |
AI Model Integration | |||
Detailed Code Comments | |||
Knowledge Graph | - | ||
Model Documentation | - | ||
Ontology | - | - | |
Source Code | - | - | |
Taxonomy | - | - |
Optional add-ons
You can add these on the next page.
Fast Delivery
+$10 - $80
Additional Revision
+$15Frequently asked questions
About Muhammad Ihtesham
AI/ML Engineer | Deep Learning | Computer Vision | RAG | Chatbots
Bannu, Pakistan - 6:12 pm local time
PEC Registered Engineer (COMP/028521) with a B.Sc. in Software Engineering and 3+ years of hands-on AI/ML development building, deploying, and shipping real production systems. I don't just build models; I deliver end-to-end AI solutions.
WHAT I BUILD FOR CLIENTS:
✅ Deep learning models for medical image analysis (cancer detection, histopathology classification, segmentation)
✅ Real-time computer vision systems using YOLOv12 object detection, ANPR, traffic enforcement, surveillance
✅ Production-grade REST APIs with FastAPI/Django containerized with Docker, deployed on AWS/Hugging Face
✅ On-device AI mobile apps using Flutter + TensorFlow Lite (published on Google Play Store)
✅ Custom CNN architectures ResNet, EfficientNet, DenseNet with transfer learning and Grad-CAM interpretability
✅ Data science pipelines — preprocessing, augmentation, class balancing, model evaluation (AUC-ROC, F1, Precision-Recall)
PROVEN RESULTS:
→ Built a live AI Traffic Violation Detection System (YOLOv12 + FastAPI + Docker) deployed on Hugging Face Spaces
→ Published "Fruit Lens" on Google Play Store on-device TensorFlow Lite fruit AI scanner with zero data collection
→ Final Year Research Project: Multi-class blood cancer classification (ALL, AML, CLL, CML) — ResNet-50 + EfficientNet-B3 + Grad-CAM, structured for publication
→ Taught Python, ML, and AI to 100+ students under the KP Government Digital Skills Programme
→ Co-founded IAR Soft published 2 production Android apps
TECH STACK:
Python (Advanced) · PyTorch · TensorFlow/Keras · YOLOv8 · OpenCV · FastAPI · Django · Docker · Kubernetes · Flutter · Dart · AWS (EC2, S3, Lambda) · Scikit-learn · NumPy · Pandas · Hugging Face · Git
I work best on:
Medical AI projects · Computer vision systems · AI-powered APIs · Mobile AI apps · Research-to-production pipelines · Data science consulting
Available for long-term contracts, project-based work, and research collaboration. Let's build something that matters.
Steps for completing your project
After purchasing the project, send requirements so Muhammad Ihtesham can start the project.
Delivery time starts when Muhammad Ihtesham receives requirements from you.
Muhammad Ihtesham works on your project following the steps below.
Revisions may occur after the delivery date.
Step 1:
Requirements review I analyse your dataset, goals, and constraints, and confirm the technical approach before starting.
Step 2:
Data preprocessing cleaning, augmentation, normalization, and class balancing to prepare your data for training.