Kseniia Arshinina

Data Scientist with 3 years of experience. Working mainly in Machine Learning (Reinforcement Learning, Neural Networks, Artificial

Intelligence).

Strong team player experienced in problem solving and building algorithms.

Together with the team created AI-based product for reducing energy consumption in buildings.

Arloid Automation

Data Scientist

June 2020 - now

• Manage the process of neural networks learning;
• Monitor the effectiveness of the learning process;
• Improve activation functions, loss functions, reward functions, etc.;

Together with the team helped reach energy savings for building owners to the level of 3-15%.

• Test and visualize the performance of the trained models;

Provided new ways of clearer explanation of the company value to its clients and helped the Sales team reduce the time of the sales cycle;

• Calibrate building simulations before training the models on them;

As the result of my work AI team got the image of how far off building simulations could be from real buildings, which helped make 35% more accurate models;

• Collect and prepare data for further BI;
• Write documentation on AI team’s workflows;

The documentation that was produced reduced time company employees needed to understand AI team’s processes (from 1 week to 1 day);

• Act as Product Manager for an internal tool for managing and monitoring learning processes. AI team now has an internal app that lets manage all learning in one app (launch, monitor, stop, keep track of all hyperparameters of each model, sort, filter, visualize, etc.). This reduced the time of working and managing the models by 100%.

Based on my performance as a Junior Data Scientist I was promoted to Data Scientist in 1 year.

Pushkin Leningrad State University

Bachelor's Degree