Zahra Toroghi

 Object Removal and Replacement Systems: Designed and implemented systems for removing and replacing objects in images and videos with up to 90% accuracy.
 Object Recognition: Developed advanced models for identifying and classifying objects in both 2D and 3D environments.
 2D and 3D Interior Architecture: Designed and optimized interior architecture using generative AI techniques, resulting in improved quality and accuracy of outcomes by
up to 90%.
 Direct Integration with Amazon: Established and managed real data contracts with Amazon, ensuring direct access to data resources for various projects.

 Developed a natural language processing model for a health trends analysis project, using classification algorithms to analyze health data. I successfully deployed this
model in a production environment, which increased the accuracy of health trend predictions by 30%.
 By utilizing Random Forest and Gradient Boosting algorithms, I optimized predictive models, enhancing their accuracy by 20%. My ability to critically assess the
effectiveness of models contributed to continuous improvement in output quality.
 In a project focused on deploying machine learning models, I utilized Kubernetes to orchestrate containerized applications. This enabled automated scaling and
management of resources, leading to a 40% reduction in deployment time. I implemented Kubernetes Helm for managing application configurations, facilitating smooth
updates and rollbacks. Additionally, I integrated Kubernetes with AWS to ensure high availability and efficient load balancing of the deployed models.

In the health trends analysis project, I developed over 15 interactive charts and dashboards using Tableau and Matplotlib, which revealed crucial patterns in health data.
This analysis improved reporting speed to management by 25%, empowering data-driven decision-making.

 In projects involving diabetes data analysis and power consumption prediction systems, I collected data from multiple sources and organized it using SQL and Python,
leading to the development of accurate predictive models that increased prediction accuracy by 25%.
 By designing and implementing automated data collection scripts using Python and libraries like BeautifulSoup and Pandas, I improved data collection efficiency by 40%,
reducing collection and preprocessing time.
 Managed the deployment of predictive models on AWS, ensuring scalability and security of data pipelines. This experience included using AWS services such as S3 for
data storage and EC2 for model deployment, leading to a 25% reduction in processing time.
 In the energy consumption forecasting project, I utilized AWS EC2 for running machine learning models and AWS S3 for data storage. Using AWS SageMaker, I trained
and deployed models, which reduced processing time by 30% and improved scalability. Additionally, I created analytical dashboards with AWS QuickSight, enhancing
data-driven decision-making.
 Developed predictive models using Python and PySpark for data analysis projects. In a project focused on power consumption prediction, I optimized data processing
tasks, resulting in a 40% increase in processing speed.
 Participated in a project where I collaborated with developers to deploy machine learning models using Kubernetes. This experience allowed me to ensure that the models
were scalable and maintained high availability in production environments.
I cleaned and validated data using Pandas and NumPy, addressing missing values and correcting inconsistencies. By performing variance tests and cross-validation, I
ensured data accuracy, boosting predictive model performance by 15%.

I developed a diabetes prediction model using the NIDDK dataset, focusing on medical indicators like glucose levels, blood pressure, and BMI. I employed classification
algorithms and validation techniques for data analysis, presenting findings through visualizations in Tableau, including scatter plots, histograms, and heat maps. This
project helped identify key trends in diabetes diagnosis and improve treatment strategies.
I presented findings through visualizations in Tableau, including scatter plots, histograms, and heat maps, which enhanced understanding of data and led to more effective
decision-making.