About Me
Hello! I'm Saman Siadati, a qualified data/AI engineer with a background in ML/AI/big data projects. I enjoy building meaningful projects and sharing knowledge through my work.
Skills
Programming Languages
- Python
- C++
ML/AI
- Scikit-Learn
- TensorFlow
- PyTorch
- MLOps
Databases
- MySQL
- MS SQL Server
- PostgreSQL
- MongoDB
Big Data
- PySpark
- Databricks
- Teradata
Data Analysis
- Pandas
- NumPy
- Matplotlib
- Seaborn
- Plotly
- Splunk
Data Pipelines
- AWS Glue
- AWS EventBridge
- Azure Synapse
Microservices
- FastAPI
- AWS Lambda
- AWS API Gateway
- AWS S3
Cloud Platforms
- AWS
- MS Azure
- GCP
BI / Visualization
- Tableau
- Power BI
Tooling
- Docker
- Poetry
- Git & GitHub
- GitLab
- Bitbucket
- Azure DevOps
Soft Skills
- Statistical Thinking
- Problem-solving
- Teamwork
Experiences
AI Trainer
Evaluates and enhances AI-generated mathematical content for accuracy, clarity, and relevance.
LLM Engineer
Recommendation system to choose the best investor for the best combination of research group and research subjects. Question answering system using the Google TAPAS LLM for table-based data.
Data / AI Engineer
Created several data pipelines for data ingestion, cleansing, manipulation, and transformation; created API endpoints, big data analytics, and data visualization with Tableau.
Data Scientist / Data Engineer
Created a risk management system that populates data files into bronze, silver and gold layer tables to extract insights.
AI/ML Engineer
Created a speech recognition and sentiment analysis tool with TensorFlow and FastAPI.
Data Engineer / Data Scientist
Created a containerized environment with OpenShift and FastAPI to extract data from Splunk and push it to an SFTP server, for use in Power BI.
ML Engineer
Generated insights from real bank transaction data (best merchant, state, city and suburb, most valuable transactions, best weekday and daytime) with exploratory data analysis, data cleansing, feature engineering and geo-spatial analysis.
AI/ML Engineer
Created an automation for mask detection in pictures and videos using TensorFlow and OpenCV.
ML Engineer
Created predictive modeling and an automated alarm mechanism for older people to reduce the risk of hospitalization and death, using anomaly detection.
Data Scientist
Web scraping for Gender Equity Victoria (women's health and prevention of violence against women) with Python, Beautiful Soup and Scrapy.
Data Scientist
Text analysis of student and staff questionnaire comments about remote technologies (n-grams, word clouds, sentiment analysis with TextBlob); EDA and visualisation of university department enrolments with Seaborn, Matplotlib and Plotly; big data analysis with Google BigQuery.
Projects
Question Answering tool with LLM
An intelligent question answering tool that combines Large Language Models with Google TAPAS, a transformer model designed for tables. It gives accurate, contextual answers to natural language queries about tabular datasets, bridging the gap between questions and complex data tables.
Recommendation system for investment in research projects
A recommendation system that identifies the best investors for each combination of research group and research subject. Deep learning models built with TensorFlow predict the most promising matches, PostgreSQL stores group, subject and investor data, and a FastAPI REST API connects the engine to user-facing applications.
Facial Expression Analysis tool
A convolutional neural network built with TensorFlow that classifies facial emotions such as happiness, sadness, anger, and surprise. A FastAPI REST API lets users upload images and receive real-time predictions, for use in marketing, gaming, and healthcare.
Mask Detection tool
An automation system that detects face masks in images and videos using TensorFlow and OpenCV. It handles multiple faces at once and classifies them by mask presence, so it can support public safety and compliance monitoring.
Service Price Recommendation system
A best-price recommendation system for computer services in Australia. Scrapy gathers pricing data from major market and sales companies, and TensorFlow models predict the best service price for the next day.
Old aged-care hospitalisation alarm system
Analysis of time-series data from sensors in home appliances to detect anomalies in daily behavior. Device signals become 24-character strings of hourly activity, and frequency metrics plus Levenshtein distance flag abnormal days without manual monitoring, enabling proactive healthcare. Predictive models are being optimised with Keras and TensorFlow (activation functions and dropout).
Breast Cancer Prediction
Prediction of breast cancer from features of digitized fine needle aspirate (FNA) images, using the Wisconsin Diagnostic Breast Cancer dataset (569 samples: 357 benign, 212 malignant; 30 features each). An optimized SVM improved diagnostic accuracy, reducing false negatives from 11 to 2.
Research & Publications
CQL is similar to SQL, not the same: Cassandra Query Language
Read moreThe first step in computational optimization: mathematical programming
Read moreGenetic algorithm: a deep dive into the natural selection method
Read moreLinear algebra: the hidden engine of machine learning
Read moreCertificates
Databricks Large Language Models: Foundation Models from the Ground Up
Databricks Large Language Models: Application through Production
Databricks Lakehouse Fundamentals
Oracle Machine Learning using Autonomous Database Specialist
Certified Artificial Intelligence Professional
Statistical Thinking and Problem Solving
Google Cloud Platform Big Data and Machine Learning Fundamentals
Advanced Data Science Specialist
Scalable Machine Learning on Big Data using Apache Spark
Deep Learning & Neural Networks with Keras
Contact
To get in touch, email samansiadati@gmail.com.