Ismat Samadov
Machine Learning Engineer & AI Systems Developer
Production ML Systems | Computer Vision | NLP | MLOps
Building scalable AI systems and production-ready ML pipelines with expertise in computer vision, natural language processing, and end-to-end model deployment across banking and healthcare sectors.
Currently developing enterprise ML solutions at Kapital Bank while creating open-source AI frameworks and contributing to low-resource language processing for Azerbaijani NLP.
Professional Experience
SQL Developer & ML Engineer
Kapital Bank
- Developing machine learning systems for fraud detection and risk assessment using SQL-based feature engineering and model deployment
- Building automated ML pipelines for transaction monitoring and anomaly detection with real-time inference capabilities
- Implementing data extraction and transformation processes for financial ML models and predictive analytics
- Creating production-ready stored procedures and database optimization for high-throughput ML feature serving
- Collaborating with data science teams to deploy enterprise ML solutions meeting Basel III and IFRS 9 compliance requirements
Business Analytics & Data Science
Unibank
- Developed predictive models and machine learning solutions for customer behavior analysis and business intelligence
- Built automated reporting systems and ML-powered dashboards for tracking key performance indicators and customer metrics
- Implemented advanced analytics using SQL and statistical methods to identify patterns in large-scale financial datasets
- Created data preprocessing pipelines and feature engineering frameworks for business intelligence and predictive modeling
- Designed A/B testing frameworks and statistical analysis methods to measure business impact of ML-driven initiatives
Risk Analyst & Data Specialist
Unibank
- Applied statistical methods and early ML techniques for credit risk assessment and loan portfolio optimization
- Developed data analysis frameworks for evaluating applicant profiles using feature engineering and risk scoring models
- Built automated reporting systems for loan performance tracking and risk metric calculation
- Implemented data quality checks and validation procedures for ML model inputs and credit decision systems
- Collaborated with fraud prevention teams on data analysis and pattern recognition for application verification
Independent ML Engineering Projects
Actively developing production-ready machine learning systems and contributing to open-source AI projects:
- Healthcare AI Systems: Built chest cancer detection and medical imaging analysis platforms
- Computer Vision: Developed tree disease detection using YOLOv8 with 93.3% mAP50 accuracy
- NLP Engineering: Created Azerbaijani language processing tools and custom NER systems
- Financial ML: Implemented probability of default models and customer lifetime value prediction engines
- MLOps Frameworks: Built intelligent agent systems and automated ML pipeline architectures
ML Engineering Expertise
ML Frameworks
Computer Vision
NLP & AI
MLOps & Infrastructure
Data Engineering
Financial ML
ML Engineering Portfolio
Production-ready machine learning systems spanning computer vision, NLP, financial modeling, and AI infrastructure. Built with modern MLOps practices and deployed at scale.
Chest Cancer Detection Platform
Medical imaging AI system using DenseNet121 with attention mechanisms for chest cancer detection. Features automated DICOM preprocessing, model serving infrastructure, and web-based diagnostic interface with explainable AI capabilities.
Tree Disease Detection System
YOLOv8-based real-time agricultural disease detection achieving 93.3% mAP50 on UAV imagery. Includes optimized inference pipeline, REST API deployment, and automated model retraining with performance monitoring.
Azerbaijani NER Engine
Custom Named Entity Recognition system for low-resource Azerbaijani language. Built with FastAPI microservices, containerized deployment, and real-time inference with custom tokenization and model fine-tuning.
Enterprise PD Scoring API
Production-grade Probability of Default system with Basel III & IFRS 9 compliance. Multi-segment ML models with sub-100ms latency, batch processing capabilities, and comprehensive monitoring dashboard.
Azerbaijani GPT Implementation
Transformer-based language model trained on Azerbaijani Wikipedia corpus. Features custom tokenization, inference optimization, and distributed training pipeline with model performance tracking.
Intelligent Agent Framework
Production-ready framework for intelligent agents with containerized deployment. Implements multiple agent types (Simple Reflex, Model-Based, Utility-Based, Q-Learning) with interactive visualization.
Real-time Fraud Detection
Streaming ML system for transaction fraud detection with adaptive learning. Features synthetic data generation, real-time feature engineering, and anomaly detection with low-latency inference.
CLV Prediction Engine
Scalable Customer Lifetime Value prediction using gradient boosting algorithms. Includes automated feature engineering, A/B testing framework, and real-time model serving with drift detection.
Interested in ML Collaboration?
I'm passionate about building production ML systems that solve real-world problems. Let's discuss how we can leverage AI to drive innovation in your organization.
Education & ML Expertise
Formal Education
Azerbaijan State University of Economics
Master of Business Administration (MBA) • Artificial Intelligence
Current
Advanced studies in AI applications for business, machine learning systems design, and enterprise AI strategy.
Mingachevir State University
Bachelor • Management of Industry
2012 - 2016
Foundation in business management with focus on process optimization and analytical thinking.
Professional Certifications
Oracle Database SQL Certified Associate
Oracle
ID: 290631207OCASQL12C
Issued: May 2022
Advanced SQL optimization and database design for ML feature engineering and data pipelines.
Machine Learning Research & Development
Machine Learning Engineering
Self-Directed & Open Source
Focus: Production ML Systems
Key Achievements:
- •Built 15+ production-ready ML systems
- •Contributed to Azerbaijani NLP research
- •Developed computer vision models with 93%+ accuracy
- •Implemented enterprise-grade MLOps pipelines
Deep Learning Specialization
Research & Implementation
Focus: Neural Networks & AI
Key Achievements:
- •Custom transformer architecture for Azerbaijani language
- •Medical imaging AI with DICOM processing
- •Multi-agent reinforcement learning systems
- •Real-time inference optimization
Core ML Engineering Competencies
Let's Build Something Amazing
Interested in ML Collaboration?
I'm passionate about leveraging machine learning to solve complex real-world problems. Whether you're looking to implement production ML systems, develop AI solutions, or discuss opportunities in computer vision, NLP, or MLOps, I'd love to connect.
What I Can Help With:
- • Production ML system design and deployment
- • Computer vision and medical imaging AI
- • NLP and language model development
- • Financial ML and regulatory compliance
- • MLOps infrastructure and automation
- • Custom AI solutions for your domain