Microservice Architecture. Machine Learning Models Inference in Real Time

Student: Krapivin Gleb

Educational Programme: Financial Technology and Data Analysis (Master)

Year of Graduation: 2020

This paper considers the use of the microservice architecture for implementation of the application, that makes real-time predictions, while using pretrained ”classic” machine learning models. The uniqueness of the solution lies in the use of the Apache Arrow Plasma – in-memory storage, which avoids the costs of deserializing data and transferring it over the network, resulting in a high response rate of microservice models, written entirely in Python. The paper describes the implementation of all the components of the proposed architecture, including: storage data loader, predictve models, third-party packages. Load testing is conducted and its results are analyzed along with the performance of the Apache Plasma compared to the the existing solutions – Redis, PostgreSQL. Keywords: Microservices, Machine Learning, Inference, Apache Arrow, In-memory storage Plasma, Python, Kubernetes, gRPC, Apache Hadoop, Kafka.

Student Theses at HSE must be completed in accordance with the University Rules and regulations specified by each educational programme.

Summaries of all theses must be published and made freely available on the HSE website.

The full text of a thesis can be published in open access on the HSE website only if the authoring student (copyright holder) agrees, or, if the thesis was written by a team of students, if all the co-authors (copyright holders) agree. After a thesis is published on the HSE website, it obtains the status of an online publication.

Student theses are objects of copyright and their use is subject to limitations in accordance with the Russian Federation’s law on intellectual property.

In the event that a thesis is quoted or otherwise used, reference to the author’s name and the source of quotation is required.

Search all student theses