Traditionally, data is collected using a dense array of sensors in every building location monitored. However, there is a multitude of issues to be considered using such an approach, including cost, maintenance, and calibration.Moreover, certain studies have demonstrated that the wrong location of the sensors may decrease the performance of a building automation system. Consequently, data should be acquired by locating the sensors in specific positions. Also, the current approach to monitoring is based on static sensing instruments deployed at specific points, which is not scalable owing to the high initial and maintenance costs associated with such sensors. In this approach, DTs perform better if the amount of data generated is vast; however, in existing assets that characterise most of the built environment, installing a significant number of sensors is not always possible. The introduction of Artificial Intelligence (AI) techniques might help understand information from a limited set of data. Implementation of virtual sensors within the building will enable a much higher resolution of data regarding IAQ in the building with minimal capital cost or disruption to the building. Combining physical and virtual sensors will result in better interventions to tackle environmental, comfort and energy performance issues in buildings. The deployment of validated data-driven (AI/ML) techniques will also result in the creation of additional metrics derived from the on-site installed sensors and meters that otherwise are difficult to be estimated.

Leading applied data analysis seminar workshops for two bachelor groups.Teaching python to the students of one of the most prestigious universities in Russia.Python [pandas, matplotlib, sklearn, numpy, scipy, Flask], SQL, a bit of JS for students' personal projects interface.Unfortunately, I was not allowed to continue teaching after I moved to Milan for my PhD.

· Researched and changed a way spark jobs build their python environments on deploy stage for all our projects.· Created spark jobs to supply our clients with fresh data (Airflow, Spark)· Implemented integration with outer systems via their APIs (Huawei ads, VK ads etc.) (FastApi, Spark, Vue.js)

Python[falcon, flask, matplotlib, numpy, pika, websockets], Docker, Rabbitmq, PostgreSQL, InfluxDB, GrafanaREST APIs using Falcon and FlaskCreated Live Data visualization module using matplotlib for complex Civil Engineering monitoring systems. GrafanaDaemons administration and creatingLinux. For example, successfully deployed all needed environment on a naked Debian machine with no internet connection, crushing it only once in processMade a service for site administrators and specialists to exchange tasks info with each other

- Hypothesis testing based on HADI framework, customer development. I made more than 1000 cold calls that resulted in 6 first clients- I also created an MVP, where clients could submit their applications for their new target ad campaign and get an ABC analysis of their sales – a hook to upsell our services- Hosted this bot on a Cloud server and established a deploy process to production

SQL, Hive, Python, AlationData Governance specialistManaged the Data Catalog based on Alation, operated the data stewardship program in whole MTS Big Data department.Implemented Data Lineage for all Data Lake replications of the company, established the same process for Data Marts, converted S2Ts to one common format.Also issued a "data lake news" bulletin, where all Data Warehouse product teams news for a month were collectedIn general, I was the MTS "Alation guy"

Internship in a civil engineering company. Mostly working in quality control department (the lab), performing qualification tests based on European standarts. Full-time.

Maintenance works on 5 square kilometers area of Eco-park, located in Tula Region, Russia. Landscape design, construction engineering.