Eyna Shabani Portrait
PIPELINE_STATUS: ONLINE

Eyna Shabani

> _

A curious and creative programmer with a deep passion for learning. I focus on projects at the intersection of medicine, finance, and education, aiming to make work in these fields easier and more engaging through artificial intelligence.

system_diagnostic_logger.py
$ python initialize_pipeline.py
Initializing artificial intelligence diagnostic pipeline...

Technical Stack & Weights

Python
Core Engine
Beginner Weight: 0.90
NumPy & Pandas
Data Struct
Beginner Weight: 0.85
Scikit-Learn
Regression
Beginner Weight: 0.85
PyTorch & TF
Deep Learning
Beginner Weight: 0.85
OpenCV
Computer Vision
Beginner Weight: 0.65
NLP
Transformer Log
Beginner Weight: 0.65
Git Versioning
VCS Standard
Beginner Weight: 0.60
SQL
Beginner
Beginner Weight: 0.30

Engineering Philosophies

Designing reliable architectures with highly optimized layers. Translating mathematical models into performant data ingestion, transformation, and automated retraining pipelines.

Adaptive Loss Optimization Algorithms
Automated Hyperparameter Calibration
Meticulous Bias Diagnostics & Auditing

Design Tools

Figma (Layout Planning) โ€” Weight: 0.30
Photoshop (Advanced editing) โ€” Weight: 0.90
PowerPoint (Presentation design) โ€” Weight: 0.60

Immersive AI Pipelines

Quantitative & Deep Learning
PRODUCTION_STABLE

AI Risk Management System

A comprehensive AI-driven system engineered to predict financial risk indicators and market volatility. Utilizes advanced LSTM layers and Self-Attention mechanisms combined with rigorous quantitative engineering to capture systemic changes.

Predictive Accuracy 94.8%
Latency Response < 18ms
PyTorch LSTM Attention Block NumPy Pandas
attention_module.py UTF-8
class Attention(nn.Module):
    def __init__(self, hidden_dim):
        super().__init__()
        self.weights = nn.Linear(hidden_dim, 1)

    def forward(self, l_outputs):
        attn_weights = F.softmax(
            self.weights(l_outputs), dim=1
        )
        context = torch.sum(
            attn_weights * l_outputs, dim=1
        )
        return context, attn_weights
Computer Vision & Medical
EVALUATION_PHASE

Medical AI Assistant

Intelligent assistant designed to aid medical specialists in detecting early-stage pathologies. Incorporates multi-layered neural network processing of medical diagnostic imaging.

Image Res 512x512
Beta Testers Active (20+)
OpenCV TensorFlow CNN ResNet-50
Reinforcement Learning
PROTOTYPE

EduML Platform

A highly customized, interactive learning framework that dynamically adapts difficulty levels and educational pathways using reinforcement learning strategies based on tracking real-time student performance.

Adaptation Rate Real-time
Decision Model Q-Learning
Q-Learning Pandas Scikit-learn Analytics Engine

Initialize Contact Node

Whether you want to collaborate on innovative research projects, discuss pipeline models, or simply talk tech stacks, my channels are open. Secure encryption handshake established.

Direct Node
eynashabani@gmail.com
Github Registry
Eyna-A
Professional Index
Eyna Shabani
guest@eyna_shabani:~