Andrei Kaliada·Jul 23From RFP to architecture: the questions I ask before designing an AI solutionWhy the models are the least stable part of an enterprise AI design
Andrei Kaliada·Feb 24How I built a Computer Vision system that understands doors on floor plansThis is a deep dive into how I used YOLO, keypoint detection, and a clever ‘virtual mirror door’ method to make everything work.A response icon2A response icon2
Andrei Kaliada·Apr 14, 2022Data synthesis for MLReal data sets are a source for creating models that will improve business efficiency and solve existing problems. Unfortunately, it is not…
Andrei Kaliada·Feb 26, 2022Generating realistic fake data using SnowfakeryIn the process of application development, fake data is often used, which is needed to demonstrate or test functionality. You can, of…
Andrei Kaliada·Dec 20, 2021Merry Christmas and YOLOv5 with object trackingWhat would you like to do during the pre-Christmas holidays? Of course, to learn something new! So I thought and decided to try myself in…
Andrei Kaliada·Dec 8, 2021Predict customer churn using Azure MLIn the last article, I did an EDA data analysis. Now, in continuation of that article, I will try to create a production-ready service. I…
Andrei Kaliada·Nov 18, 2021What can I advise a telecommunications company to improve the business?A common business problem is to understand why customers churn and what to do with it. Today I will try to investigate this for the…
Andrei Kaliada·Sep 29, 2021An affair with Data Science. How to Monetize Big DataI read an amazing book about Data Science the other day. This book is for readers who want to try their hand at data analysis and ML.