DataGo provides clients with the development of analytical infrastructure, enabling businesses to extract value from existing data immediately. This work is carried out at both the product and consulting team levels.Roles and Responsibilities:I serve as the Chief Technical Officer (CTO) and Chief Product Officer (CPO). Additionally, I led the analytics team during the first 8-9 months.As Chief Technical Officer: - Building and managing the backend and frontend development teams. - Establishing internal processes within the development team. - Managing the expectations of the development team and other related departments.As Chief Product Officer: - Building and leading the product team. - Establishing a process for collecting feedback to generate product hypotheses. - Managing the workflow related to product hypotheses.As Head of the Analytics Team: - Building the analytics team. - Establishing processes within the analytics team. - Managing the expectations of the analytics team.Key Achievements:As Chief Technical Officer: - Developed and launched the first version of the product (MVP). - Built two development teams: one for analytical products and one for self-service products. - Organized the processes for product support and new feature development. - Facilitated the transition from MVP to a more resilient and scalable tech stack.As Chief Product Officer: - Established a process for collecting product hypotheses based on research and feedback from related departments - Organized the process for handling product hypothesesAs Head of the Analytics Team: - Built a versatile analytics team capable of handling a wide range of analytical tasks for our clients. - Created conditions for the growth of analysts - Developed processes to manage analytical tasksCurrently, DataGo is self-sustaining, with product and consulting revenues being comparable in size. I am proud of this achievement!

After my experience with my own startup, I decided to take a break from managerial responsibilities and work in a line position as an analyst.As a Senior Analyst, I was responsible for the following tasks: - Metric Decomposition - Building Dashboards for Key Metrics - Conducting and Analysing A/B Test Results - Analysing the Performance of New Feature Launches without A/B TestingKey Achievements: - Created a metrics tree for the department I worked in. - Developed a dashboard in Tableau for the department. - Wrote scripts for automatic calculation of experiment results in Python (using Bootstrap, CUPED), with results loaded into Tableau. - Implemented a method for selecting control groups based on behavioral profiles to assess the results of new feature launches without A/B testing.

Jam Track is a service that separates a music file into different instrumental parts, such as vocals, bass guitar, drums, and piano. The key idea is to create backing tracks for musicians, allowing them to practice their skills or simply play along with their favorite songs and enjoy the experience.The project was developed by a team of three people.My Role: - Product Management - Marketing Management - Development Management - Financial Management - Analytics - Investments (personal savings) - Development of source separation services - Co-developed the backend of the service with a co-founderProject Development Stages: - June - August 2020: Development of MVP - August 2020: Product launch - August 2020 - January 2021: Feedback collection and product improvement: - Developed a media player - Track grouping functionality - Billing infrastructure - Marketing strategy: performance marketing, SEO - January 2021 - March 2021: Implemented marketing strategy, ensuring regular traffic from search engines and YouTube. Collected feedback. - March 2021: Decision to pivot the project (current features showed episodic use; user feedback: "The service is cool, but I only need it a few times a month"): - Shifted focus towards EdTech - Educational content: exercises, courses - Social learning: concept of reviews, where users can record their exercise performances and receive feedback from experts on the platform - March 2021 - September 2021: - Developed social learning functionality - Created a drum learning course - Created an electric guitar course - September 2021: Budget exhausted, project development frozen.The project is currently operational. You can check it out at https://jam-track.com

Key Responsibilities: - Building an analytics team. - Dashboards. - Research. - Experiments. - Developing data-driven micro services. - Establishing data infrastructure.Achievements: - Built an analytics infrastructure: YT (Yandex MapReduce) -> ClickHouse -> Grafana / Yandex DataLens. - Established processes within the team and across departments. - The team created regular, validated reports for stakeholders. - The team launched several micro services: - Client scoring for the telesales department: +20 percentage points in call connection rate; +50% in conversion to sales. - Churn Prediction: effectively predicted customer churn, though retention efforts by managers were less successful. - Launched an "Experiment Machine" for email marketing, running experiments every week. - Collaborated with Yandex's Anti-Fraud team to solve anti-fraud challenges.

Development and Implementation of a Decision Support System for Service Distribution (Forex Broker).System Components: - Dashboards for Key Metrics: CPA, CTR, Conversion Rate, LTV, ROI. - LTV Forecasting System: Predicts client revenue based on the first 7 days after registration. - Attribution Model for Registrations and Other Funnel Actions: - Detection Model for "Chinese Claimers": The goal was to predict whether a client intended to file a complaint based on their behavior. This was approached as a classification task. Challenges: - Scarcity of Positive Examples: Due to the very limited number of positive cases, we had to artificially limit the model's complexity. We effectively used gradient boosting on stumps (trees with depth 1), which allowed us to maintain the algorithm's generalisation ability and prevent overfitting.Results:Achieved a 0.8 ROC AUC (using the median value over 50 subsamples).After the model was implemented in production in August 2018, there were no complaints filed. :)

1. Dashboards for Key Metrics: - Included CPI, CPA, CTR, Conversion Rate, LTV, and ROI.2. Forecasting System: - Designed to answer the question: what will happen to a key metric if we make a specific change? Essential for planning.3. LTV Prediction System: - Predicted the revenue/searches a user would bring over a year based on the first 7 days after installing the product.4. Mobile Anti-Fraud System: - Developed a system to detect mobile fraud, successfully addressing: - Emulators - Incentivized traffic - Cookie stuffing The system utilizes heuristic rules, anomaly detection, and unsupervised machine learning.

1. Support and Development of Analytics Infrastructure and Data-Based Decision Support Systems: - Data Warehouse: Managed Amazon Redshift. - Data Sources: Included website access logs, events integrated via Snowplow, and app metrics for both the website and mobile applications.2. Development of Key Company Metrics Monitoring: - Financial Metrics - User Acquisition Metrics: Performance and email marketing. - Behavioral Metrics - Development of Onboarding Systems:3. Focused on identifying the actions users need to take to become loyal (the "aha moment"). For more details, see the blog post https://habrahabr.ru/company/lingualeo/blog/300998/

1. Analytics of Digital Fronts (B2C/B2B Websites, Personal Area, Mobile App): - Google Analytics Setup: Configured to obtain raw data for in-depth analysis. - Implemented BigQuery. - Developed a proprietary streaming system to transfer data from websites to a Hadoop cluster as a cost-effective alternative to BigQuery.Using this data, the following tasks were addressed: - Call Prediction: Predicted customer calls to the Call Center after visiting their Personal Account. - Dynamic Attribution Modeling: Built to optimise traffic acquisition channels. - Tariff and Service Prediction Model: Developed for non-logged-in users based on their visit history to segment users in advertising campaigns. - User Segmentation: Based on behavioral traits. - Conversion Factors Analysis: Identified factors increasing conversion rates for tariff and service sign- ups on the website. - Association Rule Analysis: Conducted for tariff and service connections to enhance cross-sell logic on the website. - A/B Testing Implementation: Over 40 A/B tests launched in 2 months.2. Analytics Tools Implementation for Company Websites and Mobile Apps: - Data Collection Method Development: Developed using GTM, significantly reducing development resources and time-to-market for such tasks to 1-2 days (previously took contractors 7-35 days). - Data Collection for Blocked Users: Implemented method for collecting information, accessible only on company websites.3. Managing Analytical Tasks from Other Departments: - Reduced average task completion time from 7 days to 2.5 days.

1. Customer Churn Management: - Churn Prediction Model Development: Created a model to predict which paying customers may not renew next quarter, achieving an accuracy of ~80% using Regularized Random Forest. - Churn Rules Identification and Categorization: Identified and categorized churn rules, followed by developing measures to combat churn. - IT Department Consultation: Advised on automating the model training process and the formation and categorization of churn rules.Result: Reduced churn by 2-3 percentage points after implementing the initial company lists.2. Implementation of GTM (Google Tag Manager) on Company Websites: - Reduced Time-to-Market for tasks involving the implementation of analytics codes by 100%.3. Management of Online Project Promotion Processes: - Search Engine Optimisation, PPC (Pay-Per-Click), Affiliate Marketing, and SMM (Social Media Marketing) efforts, from strategy development to execution.

Key Achievements in Digital Marketing and Analytics1. Website Traffic Generation - SEO: Implemented strategies leading to significant traffic growth. - SMM: Engaged in social media marketing to enhance brand presence. - Email Marketing: Developed campaigns that increased site engagement.Result: Increased daily unique users from 40,000 to 700,000.2. Analytics and Performance Monitoring - Yandex Metrica & Google Analytics: Collected and monitored key site metrics such as page views, session duration, return frequency, retention, churn rate, and conversions. - User Behavior Analysis: Utilized tools like click maps, link maps, and webvisor for detailed behavior studies. - A/B and Split Testing: Conducted tests using Google Experiment to refine site elements.3. Qualitative Research - Usability Testing: Conducted usability tests for website and mobile app interfaces using the "Think Aloud" method. - Advertising Content Testing: Performed in-depth interviews to assess advertising effectiveness.4. Event Registration System Implementation - Developed an internal analytical system for tracking user interactions such as page views, button clicks, and interactive elements engagement.5. User Journey Analysis - Analyzed user evolution from first visit to registration and loyalty using sequence analysis (Lichtenstein method) and hierarchical clustering (Ward method). - Identified key user paths and provided recommendations to reduce user drop-off points.6. Key Factors Influencing User Engagement - Used data from analytical platforms and applied multivariate linear regression and correlation analysis to identify key factors influencing page depth and return frequency.7. Forecasting Key Metrics - Predicted monthly and average daily traffic, page depth, visit intervals, brand search queries, and conversions.