Skip to main content

All answers

Бухгалтерская программа для sfh​

S (Situation):
At SFH Digital Marketing Agency, I was tasked with creating an automated payment monitoring system for internal accounting control. Management expected a standard solution that would provide basic monitoring and reporting functions.
T (Task):
My task was to develop this system within the standard requirements. I noticed that the requirements were quite narrow and realized that I could improve the solution by increasing the efficiency of accounting.
A (Action):
I implemented integration with banking APIs for real-time payment tracking and added features for automatic notifications of potential issues and discrepancies in payments. Additionally, I added a notification system for overdue payments at the end of the month for prepaid projects and for overdue payments after one month for postpaid projects.
R (Result):
As a result of implementing these features, accounting efficiency increased by 2.1 times, and the number of errors in financial reports decreased by 75%. Management was pleasantly surprised by the results, and the system became a key tool for financial operations control within the company.

Система отчетности sfh​

Share with me about a time you exceeded expectations
Tell me about a time when you were working on an initiative and saw an opportunity to do something bigger or better than the initial focus.

1
S (Situation):
One of the primary tasks at SFH Digital Marketing Agency was to create a reporting system for clients. In a competitive market, and with the aim of retaining clients, this system needed to automate the collection of data from various advertising platforms and present reports in a user-friendly format.
T (Task):
My task was to develop this system within the given requirements. I realized that the system could be improved by integrating various advertising platforms and providing different levels of reporting for different client tiers, which would increase the agency’s value proposition.
A (Action):
I developed a reporting system that collected data from advertising platforms such as Google Ads and Facebook Ads, processed it, and generated reports in various formats. For standard clients, reports were automatically exported to Google Docs in table form, while VIP clients received more in-depth analysis and data visualization through Power BI. The system automated data collection and aggregation, significantly improving reporting accuracy and reducing manual labor.
R (Result):
The development of this system saved approximately 2.5 hours of manual work per day for each large project. Team productivity increased by 40%, and report accuracy improved by 85%, significantly enhancing the quality of client service.
2
S (Situation)
At SFH Digital Marketing Agency, I was working on creating a reporting system for internal use. Initially, the project aimed to develop a simple system for generating reports on advertising campaigns. However, I saw an opportunity to integrate additional analytics tools that could significantly increase the system's value for the marketing department.
T (Task)
My task was to create a basic reporting system, but I proposed expanding the functionality to include real-time analytics and integration with advertising platforms for automatic data updates.
A (Action)
I developed a plan to expand the system’s functionality by adding capabilities for data collection and analysis from various advertising platforms. I also suggested integrating data visualization tools, enabling the marketing team to quickly analyze campaign results and make informed decisions.
R (Result)
Expanding the system's functionality significantly improved the marketing department’s efficiency, increasing productivity by 40% and reducing data collection time by 2.5 hours per day. Management appreciated the initiative, and the system became a key tool for data-driven decision-making.

Случай с Яндекс видео​

S (Situation):
While working at SFH Digital Marketing Agency, our agency was given an exclusive opportunity to launch a new advertising format in beta testing with Yandex. We selected one of our clients from a restaurant chain for this test, and my colleagues coordinated the budget with them.

T (Task):
My task was to develop an interface to interact with Yandex's API to launch this new ad format. I obtained the necessary access rights and documentation from Yandex to proceed with the implementation.

A (Action): On the day of the ad launch, just two hours in, I noticed massive spikes in our project's analytics. Realizing something was amiss, I investigated the API settings used for the ad configuration. All settings appeared correct, but the metrics showed a 100-fold increase in ad clicks solely due to this launch.

I discovered that the API was very raw and didn't account for running an ad campaign without manager intervention. Neither I nor our managers could promptly reach our contact at Yandex. Meanwhile, the client called us, ecstatic that some of his restaurants were filled to capacity despite the early hours.

After four hours, I finally reached the Yandex manager and explained the issue. He checked the internal statistics and informed us that we had already spent twice the budget planned for the entire month's test—we were spending $5,000 per hour. I requested an immediate halt to the ads and a thorough investigation.

It turned out that within the system, the number of ordered ad impressions was being multiplied by 1,000. Yandex initially tried to share the responsibility for the overspending with us. However, we provided all correspondence with their manager and the API documentation they had supplied. Eventually, they agreed to fully cover the expenses.

R (Result):
The client was thrilled with the overwhelming response and asked to continue the test. However, Yandex declined and returned with a corrected system and their own interaction interface a few months later. This experience taught me the importance of not engaging in test projects that lack an immediate stop mechanism and always maintaining detailed correspondence with company representatives.

Работа в ковид с Технониколь​

S (Situation)
During the COVID-19 pandemic, I worked at SFH Digital agency. At that time, we received a new project just as the lockdown began. The restrictions and legislative changes significantly impacted the project. It was large and complex, even without the pandemic, and was subject to substantial bureaucracy. Frequent changes in client-side managers, delayed start times, unclear goals, and undefined responsibilities created a lot of uncertainty.
T (Task)
My task was to organize the team's work and clearly distribute responsibilities so the project could move forward according to plan. It was necessary to quickly establish communication between the team and the client to eliminate uncertainty and start active development.
A (Action)
I initiated a series of meetings with the team and the client's management to clarify the project's goals and each participant's responsibilities. Together, we developed a project structure and assigned tasks according to each team member's expertise. I also implemented regular meetings to monitor task completion and promptly adjust the work plan if necessary.
R (Result)
As a result, clarity in responsibilities was achieved, and the team could begin actively working on the project. This allowed us to complete the first development phase 15% faster than planned, and we successfully continued to meet the client's needs.

Работа с блокировкой телеграмм​

S (Situation)
During my time at SFH Digital Agency, there was a period when Russia began blocking the Telegram messenger. Because the regulator could not locate Telegram's servers, it blocked more than 2 million Amazon and Google IP addresses.
T (Task)
We used a custom program called "Rules" to manage advertising. This program became unavailable, causing a disruption in ad management and data collection. This disruption led to overspending on advertising for our top clients, resulting in a wave of dissatisfaction. I believe that, on a global level, this was a mistake, assuming that such an event or series of events was impossible. Failure to account for this scenario led to these consequences.
A (Action)
We needed to take action on two fronts, personally calling or meeting with clients to explain the problem and provide assurances that it would not happen again. On the other hand, we had to involve ad managers in manually managing bids and controlling reports manually for the time being. We then deployed a small server for key clients, where we moved part of the "Rules" and the report generation algorithms.
R (Result)
Personal involvement played a key role in retaining clients and building trust. Despite a serious and extensive error, we did not lose a single client during this incident. This example taught me to plan actions even for the most extreme situations.

Работа над игрой​

S (Situation)
While working at SFH Digital Agency, I was involved in a project to create an online store for a gaming project. The store needed to be accessible on both mobile devices and desktops. We were close to completing the project when the client informed us that one of their key developers had left the project and they needed help developing a payment system and cross-platform purchase tracking.
T (Task)
I was very enthusiastic about this project and agreed to take on the new task. However, the task was challenging; it required creating an industrial-grade system that would protect users and prevent them from sharing their saves or passwords with others. This was my first experience working with the AES cryptographic algorithm in C#.
A (Action)
I asked my colleagues to take over some of my tasks on other projects to free up more time for this task. I also used my personal time to achieve a high level of quality for the client. Over the course of several weeks, I read documentation, explored available code on GitHub, and discussed my approaches with experts in this field.
R (Result)
Despite the fact that the game product itself was delayed several times, the implementation of the payment system and its integration with the database was not the reason for any further delays. As a result, the system I developed is still in production, and the total sales volume through this system reached $1 million over several years.

Коллтрекинг MGCOM​

S (Situation):
Before I became a developer in IT, I worked at a top digital agency in Russia, where I was responsible for promoting a real estate client. My role involved managing ad launches and overseeing data collection.
T (Task):
The client needed to gather data on incoming calls and decided that the simplest way to do this was to ask the caller, in the background, where they had heard about us. After listening to dozens of calls, I realized that this method was inaccurately reflecting the results of our advertising efforts.
A (Action):
I approached my team leader, and we scheduled a call with the client to propose implementing a call tracking service to accurately measure the source of calls. Initially, the client was hesitant, as it required significant upfront costs. I suggested running the system in a test mode for just one property. After some discussion, we agreed on this approach with both the client and the call tracking service provider.
R (Result):
After several months of testing, the system provided accurate statistics on the performance of ads and other channels driving traffic to the website. We were able to improve the accuracy of collected data by more than two times. This system also allowed me to optimize the advertising campaigns effectively moving forward.

Ковид и проект доставка​

S (Situation):
When COVID-19 hit, my agency, SFH, and I had to quickly adjust our processes. At the start of the lockdown, it seemed like everything was going wrong. Half of our clients were from sectors like malls, restaurants, and fitness centers, and all of them stopped their services overnight. One of our clients in the restaurant business lost almost their entire development team, as they sought more stable job opportunities. The remaining team decided that the best way to save the business was to create their own delivery system to avoid paying high commissions to third-party platforms like Uber Eats, especially as the business was on the brink of closure.
T (Task):
My task was to develop a delivery website that would integrate with the daily menu of each restaurant. The project needed to be executed rapidly, with features being added on the fly, all while working under extreme time pressure.
A (Action):
Tight deadlines: I developed the platform in emergency mode, without time for proper planning. Features were added during production, and the platform was launched just two weeks after development began.
Lack of testing: Due to the time constraints, there was no opportunity for full-scale testing. I had to fix issues as they appeared in real-time, which posed a risk to system stability, but this was the only option in such a crisis.
Hosting on a local server: We couldn’t immediately deploy the app to the cloud due to the financial limitations caused by the crisis. I initially deployed the system on the client’s local server, which increased the risk of failure and scalability issues. Only after business processes stabilized were we able to migrate the system to the cloud for better reliability and availability.
R (Result):
Despite the challenges, the application helped the business survive the crisis. It retained a significant portion of the client base, ensuring a stable flow of orders and leading to a rise in online sales. Three months after the launch, the in-house delivery system accounted for 30% of all restaurant orders, reducing reliance on third-party platforms and cutting costs by more than 25%.

История про то, как посетил Турандот и нашел новых клиентов​

S (Situation):
In early 2019, before there was any hint of a global pandemic, I was working at the SFH agency. I, the agency's managers, and the leadership of one of our clients in the restaurant business were having a meeting at the client's restaurant.
T (Task):
The topic of our meeting was the development of the largest project for our client, part of a restaurant conglomerate. The meeting itself took place at this restaurant. The task was somewhat vague for me because I wasn't acting as a marketer but rather attending as a technical specialist.
A (Action):
After the meeting concluded, I stayed at the restaurant to have lunch. During that time, I observed an interesting situation — an influx of many clients from China. It turned out that several tourist groups were visiting the restaurant. I suggested creating a small webpage for Chinese tour companies to attract customers from that market.
The restaurant itself found a contractor to host the website on their platform.
R (Result):
This idea allowed the client to create a new customer segment — tourist groups — and start accounting for them in their sales. Over the course of a year, this segment accounted for 20% of all visitors.

Как я работал с JWT токенами - DVE​

S (Situation):
At my last job at Digital Vertex Enterprises, my team and I worked on a CRM project for a newly funded startup.
T (Task):
I was responsible for developing a robust authorization system and implementing a permissions management framework to secure client-server interactions. The primary task was to ensure that the system could securely handle sensitive user data and that only authorized personnel could access specific resources and data.
A (Action):
To address this challenge, I first designed and implemented an authorization system using JWT tokens. This allowed for stateless authentication, which improved the scalability and security of the application. I also developed a detailed client-server permissions system that included role-based access control (RBAC). This system defined permissions based on user roles, ensuring that only authorized users could access particular resources and perform specific actions within the CRM.
R (Result):
As a result, the robust authentication and permissions systems I developed provided reliable protection of sensitive data, and user authentication was both efficient and secure.

Проект Кейтеринг​

S (Situation):
At SFH Digital Agency, I was responsible for developing landing pages for one of our clients, who was part of a conglomerate. Each director in this conglomerate acted as an independent entity and contractor. The client was very demanding and frequently changed their requests, which created tension within the team and jeopardized the project's deadlines and budget.
T (Task):
My task was to implement the client's advertising campaigns on time, manage changing A/B tests for the landing pages, and ensure their successful execution despite frequent changes in the technical specifications. I understood that it was crucial to focus the client's attention on the strategy of achieving sufficient results, as without enough traffic, the testing could fail.
A (Action):
Understanding this, I worked diligently to earn the client's trust and convince them of the correctness of our approach. I held several additional meetings, both online and in person, to persuade the client not to change the strategy until the necessary amount of data was obtained to make decisions about the results of the advertising and testing efforts. For this client, it was the first experience of outsourcing their projects, which caused discomfort and a desire for more control over the situation. Realizing this, I also set up automated reporting that the client received via messenger. This helped meet their requests, reduced pressure on me and the team members, and eventually turned them into a strong supporter of our approach. We developed a unified language of communication.
R (Result):
As a result, we observed a conversion rate increase every six months with test cycles ranging from one to two months, with conversion rate improvements ranging from 5% to 50%. However, looking back, I realize that many problems could have been avoided if we had conducted more detailed inquiries about the client's past experiences at the outset. Moving forward, we developed a separate approach for similar clients, allowing us to meet their expectations faster through more thorough work even before the project's start.

MGCom и гиперсегментация​

S (Situation):
When I first started working in digital advertising at a top digital agency in Russia, I was tasked with launching ads for a client in the car leasing industry.
T (Task):
I suggested trying a new advertising method called hyper-segmentation. This approach involves creating a unique image background and headline tailored to each user query. However, my team leader rejected the idea, citing the need to launch the ads quickly and within the client's budget, which might not support such a test.
A (Action):
I persisted and convinced the team to plan a pilot test using this approach within one of the ad campaigns. The hyper-segmentation method was tested on a small scale to gauge its effectiveness without overextending the budget.
R (Result):
The test resulted in a 15% increase in conversion rates compared to the ad campaigns without hyper-segmentation. Later, when I transitioned into development, I implemented an interface for quickly launching hyper-segmented pages for premium clients, further optimizing the process

Я получил 2 разных приоритета - история про Правила и новые запуски​

S (Situation)
While working at SFH Digital Marketing Agency, I encountered a situation where two managers requested urgent tasks that required significant resources and time, but which conflicted with each other in terms of priority. One manager requested an immediate update to the client interface, while the other insisted on urgent optimization of the server infrastructure.
T (Task)
I needed to determine which task was more critical to the company overall and decide on the order of execution to minimize risks and ensure both managers were satisfied.
A (Action)
I arranged meetings with both managers to discuss their requests and clarify which tasks were critical to the business. After analyzing the business priorities and the impact of each task on the company’s overall goals, I proposed a phased approach where the urgent infrastructure optimization was completed first, followed immediately by the interface update. I also coordinated resources and timelines with both parties to ensure clear understanding of the task execution order.
R (Result)
As a result, both tasks were completed on time, avoiding any disruptions and ensuring high satisfaction for both managers and end-users. The infrastructure optimization led to a 20% reduction in server response time, and the interface update improved user experience, resulting in a 15% increase in customer retention.

Приоритеты при большом количестве клиентов - матрица Эйзенхауэра​

S (Situation)
During my time at SFH Digital Agency, I simultaneously managed multiple projects and participated in client negotiations. Each client had unique needs and expectations.

T (Task)
My task was to demonstrate to clients that they were always at the center of our attention. To achieve this, we often created initial prototypes even before signing contracts, showing that we were already thinking about their needs and working to solve their problems. This helped us bypass bureaucratic hurdles and build trust with the client from the start.

A (Action)
To manage multiple projects, I used the Eisenhower matrix to prioritize tasks. I did this every week at the beginning of the week, which helped me form my client task pool. I kept clients informed about the progress of tasks to ensure that we could quickly adapt to any changes in their requirements. If any tasks fell behind schedule or required revisions, our team put in extra effort to complete them on time. We also analyzed each case of revision and sought ways to automate the process that caused the need for revisions.

R (Result)
This strategy allowed us to show clients that we were attentive to their needs, demonstrated our deep understanding of each project, and were ready to respond quickly to changes.

Как приоритеты задач в DVE помогли мне закончить проект раньше времени​

S (Situation)
While working at DigitalVertex Enterprises, I frequently encountered multiple tasks that needed to be completed simultaneously, especially during the development of a project management system for a real estate company. The projects varied from optimizing the user interface to configuring server infrastructure, each critically important to the client.
T (Task)
My task was to effectively manage these tasks to ensure their timely completion and maximum value for the client, all while maintaining deadlines and work quality.
A (Action)
I developed an approach to task prioritization based on their impact on key project metrics, urgency, and resource availability. I prioritized tasks that directly impacted final deadlines and were critical to the project's success. Next, I considered available resources and the possibility of parallel task execution. I used Agile methodology, which allowed me to quickly reassess priorities and adjust the plan as requirements changed.
R (Result)
This approach enabled me to complete the project management system two months ahead of schedule, improving system performance by 35%. The client was satisfied with the result, leading to a new contract for continued collaboration.

История про оверинжиниринг​

S (Situation):
When I first started learning programming, I mostly copied existing solutions. In 2017, I came across a book on design patterns, which piqued my interest. Although I wasn’t yet working as a developer, I joined my colleagues' small pet projects at work to review their code and understand the structure. One of these projects likely involved parsing prices from websites and comparing competitors' prices.
T (Task):
I was eager to apply my newfound knowledge of design patterns, so I decided to implement a Multiton Singleton pattern in one of the projects. I wanted to use this pattern to demonstrate my understanding of abstractions and design principles.
A (Action):
I wrote my own solution using the Multiton Singleton pattern but quickly realized that I had overcomplicated the code. When I shared the code with my colleagues, they pointed out that I had over-engineered the solution and advised me to strengthen my understanding of OOP principles.
R (Result):
While my solution didn't break the project, it was not adopted due to its complexity. This experience taught me a valuable lesson: mastering a tool doesn’t mean applying it everywhere, and over-engineering can be detrimental to a project. I learned to approach each problem with a more pragmatic mindset.

Мое хобби в играх​

S (Situation):
I was working on my own 2D game in C# with Unity and wanted to implement an inventory system with drag-and-drop functionality, similar to what is seen in Diablo 2. However, I encountered several issues with getting the drag-and-drop mechanic to function properly. The system wasn’t behaving as expected when I tried to drag items from one slot to another, and the objects wouldn't snap correctly into place. This was critical for the game’s overall functionality and user experience.

T (Task):
My goal was to develop a fully functional drag-and-drop inventory system that would allow players to interact with their inventory items seamlessly. The system needed to handle different item types, positions, and snapping behavior while maintaining a smooth user experience. I had to figure out how to implement this feature correctly and efficiently, as it was an essential component of the game.

A (Action):
To solve this, I took two weeks off from work to fully dedicate my time to research and development. I explored Unity’s built-in UI tools, such as the Event System and Canvas Groups, which are commonly used for drag-and-drop mechanics. I also evaluated several third-party frameworks to see if any of them could simplify the process.
I created multiple prototypes, testing different approaches, including manually coding the drag behavior and experimenting with colliders and event triggers. I consulted forums, watched tutorials, and reviewed GitHub repositories with similar implementations. Gradually, I learned how to handle the snapping logic and item placements, as well as how to ensure the system was flexible enough for future game updates.

R (Result):
After extensive research and trial-and-error, I successfully implemented a working drag-and-drop inventory system that allowed players to drag items between inventory slots, with items snapping into place just like in Diablo 2. The system was intuitive, flexible, and enhanced the overall gameplay experience. This achievement not only improved my game, but also expanded my knowledge of Unity’s UI system and C# event handling.

История про разницу культур в работе​

S (Situation):
When I transitioned from working in a company based in Russia to Digital Vertex Enterprise in the US, I encountered some challenges adapting to the different communication styles and cultural expectations in the workplace. While both companies used similar methodologies like Agile and Scrum, the way feedback and collaboration were approached was noticeably different. In Russia, feedback tends to be more direct, whereas in the US, a more diplomatic and structured form of communication was preferred.
T (Task):
My task was to adapt to the new cultural norms of communication in the US while maintaining effective teamwork and collaboration. I needed to provide feedback and express my opinions clearly, without coming across as rude or blunt, but without losing my professional voice. It was important to ensure that my feedback would be well-received and not misinterpreted.
A (Action):
To address this, I invested time in learning how to use the Sandwich Method, which is a way of delivering feedback by balancing both positive and constructive remarks. I practiced providing feedback in this manner—starting with a positive comment, addressing the area of improvement, and concluding with encouragement. Throughout this adaptation process, I made sure that I was never rude or blunt. I remained professional and considerate, ensuring that my directness did not come off as harsh but was instead framed in a constructive way that aligned with the team’s communication style.
I also engaged in open discussions with my colleagues to better understand their expectations and how I could adapt my communication to fit the team dynamic. I paid close attention to how feedback was given and received, learning from those interactions to improve my approach.
R (Result):
By adapting my communication style, I was able to integrate into the team more smoothly. My feedback was received positively, and I was able to build strong working relationships with my colleagues. This adjustment in communication style allowed me to express my ideas clearly without any misunderstandings, and I became an effective contributor to the team. In the end, this approach helped improve overall collaboration and fostered a productive and respectful work environment.

История про сложную тему, как ребенку - lambda​

S (Situation):
At Digital Vertex Enterprises, I was working on the development of an Analytics and Ad Management System for a major retailer on Amazon. A new manager with no technical background joined the team, and I needed to explain how our data collection process worked. We were gathering data from various advertising platforms like Amazon, Google, and Meta, and the Amazon API posed specific challenges since we had to use two different APIs (store and advertising). The advertising API was incomplete, missing some campaign types, and some data still had to be fetched manually.
T (Task):
My task was to explain the entire data collection process to the manager in simple terms, as if I were explaining it to a 10-year-old. This included explaining how we gather data through Amazon APIs, store the data in the cloud using Amazon RDS, and use AWS Lambda and EC2 servers for automation and processing. The goal was for the manager to understand the system well enough to participate in decision-making without getting lost in technical details.
A (Action):
I decided to use analogies to simplify the explanation:
I told the manager that the Amazon API is like two different books: one book talks about products (the store API), and the other talks about advertising (the advertising API), but the advertising book is missing some chapters (data). So, we have to get some information manually (through requests).
AWS Lambda is like an alarm clock that goes off at certain times and automatically starts a program to collect data. Think of it as an automatic vacuum cleaner that starts at the same time every day and gathers information from Amazon, Google, and Meta.
We use cloud storage in Amazon RDS (like a giant invisible closet) to store all the collected data. This "closet" is available anytime, and we can pull out information whenever we need it.
Our server, which processes the data, is like a powerful computer that runs in the cloud (EC2), managing all the tasks related to data processing.
These analogies helped simplify the complex process and made it easier for the manager to grasp.
R (Result):
As a result, the manager not only understood how the system worked but also became more confident in asking questions and participating in discussions. This improved communication between the technical and business teams, as the manager now understood how the data collection process influenced decision-making and ad management. My efforts to explain the system also helped reduce decision-making time, as the manager was no longer confused by technical jargon.

История про приют для животных​

S (Situation):
This story isn't related to a business setting, but rather to a period in my life when I dedicated all of my free time to volunteering at an animal shelter. It all began when my dog and I encountered two stray dogs during one of our walks. In Russia, there wasn't a centralized system for capturing stray animals, so these dogs lived freely on the streets.
A few months after our first meeting, I noticed that both dogs were pregnant. One of them gave birth to 13 puppies near a railroad, which posed a serious danger to them. I contacted social services, and they helped relocate all the dogs to a nearby animal shelter.
T (Task):
From that point forward, my main task was to find homes for all of these dogs. The window of opportunity for rehoming puppies is best before they reach six months of age. I knew I had to act quickly to help them find families and avoid them growing up in the shelter.
A (Action):
I visited the shelter every day to help care for the dogs—cleaning up after them and feeding them. After a few weeks, I decided to take a more proactive approach to increase their chances of adoption. I created an Instagram account for the dogs, organized professional photoshoots, and even built a simple website using a website builder to showcase each dog's story and personality.
I leveraged social media and networking to spread the word about the dogs, reaching out to potential adopters both locally and internationally.
R (Result):
After six months of dedicated effort, all 13 dogs were successfully rehomed, both within Russia and abroad. It was incredibly rewarding to see each one of them find a loving home and to know that my efforts made a significant impact in their lives.

Яндекс видео, как YouTube последовательности​

S (Situation)
While working at SFH Digital Agency, my colleagues were running ad campaigns on YouTube, where YouTube had recently introduced a new feature called sequential videos. This feature allowed advertisers to create scenarios based on whether a user watched a certain percentage of an ad. Understanding the value of this feature, I decided to build a similar system for Yandex, which was more important for the Russian market as a search engine and video platform but lacked this capability.

T (Task)
My task was to create a system for Yandex that could replicate YouTube's sequential video functionality, allowing us to group viewers into cohorts based on how much of a video they watched. This would enable our clients to better target their audience by creating specific advertising sequences for different viewer segments, optimizing ad performance.

A (Action)
To implement this, I developed a system that:

  • Created cohorts for each video based on its unique ID.
  • Required each video to gather 1,000 unique views and to run for a minimum of 2 weeks to collect sufficient data.
  • The data was then segmented by the percentage of the video watched (e.g., 25%, 50%, 75%, 100%).
  • Each segment followed a sequential advertising script, similar to YouTube's feature, allowing targeted follow-up ads based on how much of the video the user watched.

We tested this solution on a project selling tickets for a New Year’s corporate event. In Russia, New Year is celebrated more widely than Christmas, and the ticket price for the event was $5,000 per person.

R (Result)
The results were impressive—thanks to more precise audience segmentation and the use of sequential advertising scenarios, all tickets were sold out before December 1st. The system I built enabled highly effective targeting, driving significant sales for the event.

Как делали АБ тесты​

At SFH Digital Marketing Agency, I was responsible for preparing and executing numerous A/B tests as part of our marketing strategies.

S (Situation)
My role involved setting up analytics, configuring alerts for unforeseen issues, and planning A/B tests together with our manager. We needed a system to better predict how long an A/B test would take based on the client’s budget, the average cost per click (CPC), or the desired number of visitors. The goal was to ensure that the tests would be effective and resource-efficient.

T (Task)
My task was to create a simple prediction system that could estimate the time required for each A/B test by taking into account the client’s budget, the average CPC, or the target number of visitors. The system also needed to suggest whether to increase the budget or abandon a test if the initial inputs indicated that the test wouldn’t generate enough results to be meaningful.

A (Action)
To achieve this, I built a system that analyzed user behavior on the website using Yandex's behavioral analytics. This allowed me to propose meaningful changes based on deep user insights, rather than basic suggestions like changing button colors. I incorporated the following:

  • The system could estimate the required time for each test based on input data like budget, average CPC, and traffic goals.
  • It provided recommendations to either adjust the budget or drop the test if the expected results were not achievable with the given parameters.
  • I also developed a test evaluation scale, which measured test sufficiency based on traffic volume, traffic distribution, and the test duration in days.
  • The final test results were presented in an aggregated format, showing both planned and actual outcomes based on key metrics.

R (Result)
Though some tests resulted in no significant changes, most of our tests led to an increase in conversion rates between 5% and 50% in the target actions. The system improved our ability to predict the success of A/B tests, which in turn led to more efficient use of the client’s budget and better overall performance in our campaigns.

История про обучение и платформу для этого​

At SFH Digital Marketing Agency, the marketing specialists underwent necessary certifications from Yandex, Google, and Meta.

S (Situation)
After talking with my colleagues, I realized that we needed to create our own knowledge base and training platform. My stakeholders wanted to ensure that when specialists communicated with clients, the clients would be 100% confident that the specialist wasn’t a generalist, but someone who had expertise in the client’s specific industry.

T (Task)
My task was to develop a customized training platform that would address our specific needs. The platform needed to provide clear learning paths and ensure that each specialist could progress through well-defined stages of training. This would not only improve our team's expertise but also build trust with our clients.

A (Action)
To achieve this, I:

  • Interviewed managers to understand the key learning segments required and how the training roadmap should look.
  • Researched other training platforms to identify best practices, focusing on effective ways to deliver content and assess comprehension through quizzes and tests.
  • Developed a platform where managers could onboard new employees. The platform clearly outlined the steps a specialist needed to follow to progress from one level of expertise to the next.

R (Result)
As a result, the platform provided a structured learning path for each new employee. Specialists had a clear understanding of what they needed to do in order to grow and advance to the next level. This ensured that our team was not only certified by external platforms but also deeply knowledgeable about specific industries, which gave clients greater confidence in our expertise.

DataDriven подход - программа Правила​

At SFH Digital Marketing Agency, we needed to develop a system to manage advertising campaigns through a unified interface.

S (Situation)
As a company, we faced the challenge of managing numerous advertising campaigns across multiple platforms. The process was fragmented, requiring the use of several different programs. We needed a system that allowed campaign uploading and configuration in a single window. The system had to enable bid management at the keyword level and allow the setting of schedules for campaigns.

T (Task)
My task was to develop a system that simplified campaign management by consolidating everything into one interface. The goal was to make the process easier for our marketing managers by providing a unified system for creating, managing, and optimizing ads across multiple projects, ensuring seamless operation and control.

A (Action)
To get started, I:

  • Spoke with managers to understand their pain points and requirements. Drawing from my own past experience in marketing, I explained how I would want to use such a system if I were a marketing manager.
  • Created a clear plan with specific requirements for the early version of the program, ensuring that it met the needs of the team, especially in terms of managing bids and scheduling campaigns.
  • Began the development of the program, focusing on creating an intuitive and efficient interface that allowed for easy campaign setup and management.

R (Result)
The program became an integral part of the agency's workflow and was incorporated into all checklists for setting up and managing advertising. Over the course of 4 years, it processed all 100 projects within the agency, managing thousands of campaigns and hundreds of thousands of settings and targeting adjustments. Overall, the system helped optimize advertising spend, leading to savings and increased efficiency, with a total ad optimization of over $1 million.

Как ты пришел в IT - MGCom & Google​

I started developing small websites back in school, but my deep interest in the field emerged later. This is an interesting aspect of my background.
In 2016, I worked at one of the top digital advertising agencies. One of our clients for ad placements was Google, and agency employees often visited the company’s office for meetings with staff, gaining knowledge about Google’s beta products, and so on.
At one of these meetings, a Google representative mentioned that the company believed the future would be driven by small, custom scripts designed to optimize ad performance. After returning from that meeting, I delved into the topic of scripting and wrote my first solutions. Since then, I’ve been increasingly immersed in this world, starting with automation and then moving into front-end development.
During the COVID period, I also took up game development in C# as a hobby and participated in several game jams, which are similar to hackathons.
Now, I am fully engaged in full-stack development and plan to continue in this direction, with a detailed plan for the coming years.

Почему нужно нанять тебя?​

Several factors set me apart from other candidates.
One of my key strengths is my ability to solve business problems. I constantly ask myself - what can I do better? Do we have code that needs documenting? Is there any code that needs to be fixed?
Another crucial strength is my methodical approach to problem-solving, breaking down complex issues into manageable tasks to create efficient and elegant solutions.
The third significant aspect is collaboration. I excel in a team environment, actively interacting with colleagues to understand the project requirements, share ideas, and collectively achieve the project's goals.
To sum up, my diverse skill set, ability to solve problems, focus on code quality, collaborative approach, and commitment to continuous learning make me an exceptional candidate for a developer role.

Кем ты видишь себя через 5 лет​

In the next 5 years, I see myself as a full-stack developer, continuing to work with React and TypeScript. However, I also want to learn Rust. I plan to use AI platforms to develop advanced applications.
To achieve this goal, my strategic plan includes a deep dive into the full development cycle and a more detailed study of software itself.
Overall, my main objective is to be at the forefront of AI-driven development, contributing to the creation of innovative solutions and collaborating with teams to push the boundaries of technology.

Опиши свой обычный день​

I like to start my workday with task planning. For this, I prefer to be at my workstation 15 minutes before the workday begins.
First thing, we have a call with the team. In this brief meeting, we review the task stack in JIRA and the progress throughout the sprint. Our goal is to complete most of the priority tasks. To determine priorities at the start of the sprint, we distribute tasks according to the Eisenhower matrix principle.
If any team member is having issues with their tasks, we can organize an additional call to collectively solve the problem. Personally, I like working with the Pomodoro technique. My day is divided into meaningful blocks with short breaks in between. This allows me to dive deep into tasks, read others' code, spend adequate time on documentation, and even keep up with news or interesting articles in our field.
I have developed a habit of summarizing my day at the end and writing a brief note about what was done and what I need to do tomorrow or discuss on the call for each task I'm involved in.