BeeTech 2021: Review of Big Data Reports, Artificial Intelligence, IT Architecture, QA, Back-End

Yesterday I posted the first part of the video from the BeeTech conference that we held in April. Today – reports from the two remaining streams. Here, from building a QA department from scratch and changing the IT architecture to DIY solutions in system integration and developing a prototype of a military robot using computer vision.

Engineering stream

Topics for engineers, software and database developers, and web designers were presented here.

QA Beeline: a service approach. How to build QA from scratch in a large company, Oleg Kuts, Head of Product Quality Assurance

Oleg talked about how the processes in our company are built, the difficulties we faced when building QA, and plans for the future. He also shared his vision of how to build processes when working with a distributed team. Warning: there are many cats in the presentation, which immediately captivated the conference participants.

DevOps at Beeline Kazakhstan, Alexey Fateev, DevOps engineer

Alexey’s presentation was devoted to how DevOps processes are organized in our company, as well as a story about a single standard, infrastructure as a code in GitLab and continuous code delivery.

Early testing: how to save company money by improving quality, Zumrat Akhmetova, Middle QA Engineer at Kolesa Group

Zumrat shared the benefits of early testing and gave advice on how to implement it.

The Internet of Things at the Service of System Integration. And what has to do with programming, Viktor Khomenko, expert on the development of E2E solutions Beeline Kazakhstan

Victor introduced us to several interesting projects that can be closed with DIY solutions. If you like non-standard solutions to complex and non-trivial everyday tasks, be sure to watch the video.

What the bugs are talking about in production – clearly and in confidence, Alexandra Laikhtman, QA Lead in the Samokat.ru team

Alexandra talked about her phobia of bugs in production and how she works with it. This is one of the most interesting and emotional talks according to the audience.

IT architecture at Beeline Kazakhstan, Evgeniy Kim, senior expert on IT architecture

Evgeny’s report describes in detail the company’s transition from Telecom to Digital. Evgeny told how the IT architecture is changing along this path and why a dedicated architect is needed for such tasks.

Practices versus Habits: Optimizing the Development Process, Nurlan Abyken, COO at Sprint Squads

Discussion report on whether it is necessary to optimize processes or work more efficiently according to a familiar and well-established scenario.

At the end of the stream panel discussion Development culture: what it is and how to live with it. The moderator was Binali Rustamov from Beeline Kazakhstan, and the participants were Nikolay Kindyakov (Kolesa Group), Madiyar Nogaev (Beeline Kazakhstan) and Artyom Shalhakov (Global Talent).

Big Data Stream

The reports of this stream will be of interest to data scientists, analysts and data management specialists (your CEP).

How we live with Apache Airflow, Syrym Sydykov, head of Big Data development service at Beeline Kazakhstan

From Syrym’s speech, you will learn about our experience in implementing Apache Airflow and the stages of tool development.

Decoding binary files in Apache Spark, Aitimov Anuar, Tech Lead in Big Data development service in Beeline Kazakhstan

Using the example of one of our projects, this applied report introduced the conference participants to the problems of performance optimization, reducing memory consumption in Apache Spark, and thrifty use of resources when writing code.

How we solved the problem of text recognition in pictures, Sayan Omarov, Big Data analyst at Beeline Kazakhstan

We discussed the relevance of this task and considered the general pipeline of work on text recognition in pictures using one of the cases as an example. Based on the report, we even wrote an article on Habr

Using Computer Vision to Build a Military Robot Prototype, Igor Littig, Computer Vision Engineer Verigram

Igor presented the case of creating a working prototype of a military robot using computer vision. From the video you will learn why we did it, how we coped with the task and how we analyzed the results.

Computer Vision Problems in Industry, Alexander Belugin, AI Application Expert in Industry

The report is devoted to the current directions of technology development on the example of using neural networks to solve business problems, including about the technology of “computer vision” in mining equipment.

We are developing a credit conveyor from scratch in Beeline Kazakhstan, Dair Mustakhimov, Big Data analyst

Dair’s applied report made it possible, using the example of one of our projects, to show how boosting works, to make a comparison with other ML algorithms, to explain what the model is for and what product we are developing.

We are building our own data platform: from reports and triggers to complex pipelines, Andrey Markin, Data Platform Manager at Samokat.ru

Scooter Business it is the creation of a new level of urban infrastructure for the delivery of goods based on its own data platform. Andrey spoke about the functional structure of the platform, what data services we use, what logic is behind this whole structure.

Panel discussion in this stream “NLP-implementation nuances: why business needs it and development prospects”. Leading Andrey Ostafichuk, Chief Data Officer at Beeline Kazakhstan.

Discussion participants: Uakhat Bastimiev (COO, co-founder, Verigram), Ivan Yamshchikov (Associate Professor at the Higher School of Economics in St. Petersburg, Head of the Natural Language Laboratory of Yandex LEYA and the Higher School of Economics, author and host podcast “Let’s Get Out!”) and Alexander Pak (IICT Head of Laboratory Intellectual Analysis of Big Data 2018-Currently, FIT KBTU Assoc. Prof 2018-Currently).

Discussed NLP trends in 2021 in Kazakhstan and in the world; argued about whether it is worth investing time and effort in voice assistants in the Kazakh language and who needs them more – business or users.

This was a quick recap of the online conference presentations. All streams are collected in YouTube playlists, if you want to watch all videos at once, it will be more convenient for you to do this. Stream “Management” can look here… Engineering Stream Playlist hereHere collected all videos about Big Data.

Happy viewing!

Similar Posts

Leave a Reply

Your email address will not be published. Required fields are marked *