Artificial Intelligence at SIT: A perspective of an expert
Andrey Sharapov is a data scientist and data engineer at Schwarz IT. He is currently working on various projects related to machine learning and data product development. Previously, he spent 2 years at Xaxis where he helps to develop a campaign optimization tool for GroupM agencies, then at TeamViewer, where he led data science initiatives and developed a tool for customer analytics. Andrey is interested in „explainable AI“ and is passionate about making machine learning accessible to the general public.
What are your main topics as a data scientist at SIT? Which use cases are you currently working on?
I have very wide responsibilities. Primarily, I am working on machine learning bases product that helps promotion managers to make better decisions. This involves algorithm development, system architecture, the industrialization of the product, etc. Additionally, I work together with the infrastructure team to bring new technology that can be used for Data Product development.
What algorithms and tools are you using?
We heavily use Python and Machine learning algorithms available for it such as ones in sklearn. We also use Keras for some problems. Also, Hadoop, Spark, and Kubernetes are a part of our technology stack.
Not always things are working out: according to your experience how high is the level of success for data projects and which success factors are the most relevant?
We try to make all our projects succeed. The major challenges are the scaling and robustification of the software we write for all countries. Very hard to make everyone happy. Another major challenge is communication with final users. These people rarely become a part of the product feature establishment.
The retail sector is considered an industrial section with a high potential for data science and AI, would you agree with this statement? What else do you think will be relevant in 2020/2021?
Retail has a high potential for AI. For us, the major challenge is product industrialization.
Dr. Andrey Sharapov präsentiert am 13. November 2019, dem ersten Tag der Data Leader Days 2019, über „ Building a bridge between business & data science using Explainable AI“.
Tickets für die Konferenz finden Sie auf der offiziellen Startseite, unter www.dataleaderdays.com.