Community

• Collected and preprocessed data to build a time prediction model that improved the accuracy of rendering job estimates.

• Designed and implemented a visual predictive model using synthetic data for internal research.

• Built a sentiment analysis tool to assess user feedback and created a recommendation system based on user behaviour patterns.

• Trained a computer vision model to detect vehicle scratches and damage from images.

• Built a containerised service using Docker for model deployment.

• Enhanced image preprocessing pipeline to boost model accuracy and generalisation.

• Built and maintained backend modules for a rendering SaaS using Python.

• Developed a chatbot to automate communication between technical support and clients.

• Managed and optimised MariaDB and PostgreSQL databases for internal tools.

• Trained and deployed face detection and feature extraction models to production.

• Created APIs and inference pipelines using Triton Inference Server and TensorRT.

• Designed a video stream processing system handling input from multiple cameras.

• Increased vector search performance 20x by optimising FAISS-based similarity search.

• Built dockerised infrastructure for scalable deployment.

• Led development of face and license plate analytics across 4 engineering teams.

• Built ML models for driver identification using sensor data (RF, MLP, DT, KNN).

• Researched video-based action recognition (publication under review).

• Created a facial recognition API using FastAPI and OpenVINO Model Server.

• Supported academic research with practical deployments and experiments

• Led more than 5 research and development projects focused on building human-machine interfaces to control robotic systems via Android apps, WinForms applications, and console tools.

• Conducted research in robotic control and trajectory planning, contributing to academic advancements and real-world implementations.

• Worked on sensor fusion tasks involving PIR, sonar, camera, touch, and gyroscope inputs — integrating and processing data from multiple sources.

• Completed two projects using microcontrollers and single-board computers such as Arduino, PIC32, Raspberry Pi, and Jetson TK1.

• Delivered two working computer vision applications involving object recognition, emotion detection, and a “leader-following” system (robot tracking a moving subject).

• Built a cross-platform mobile app that offers smart outfit recommendations and wardrobe management features.

• Implemented outfit creation tools with a clean, mobile-first UX using React Native and Expo.

• Connected frontend to backend using Supabase and Prisma for rapid, scalable development.

• Designed intuitive logic to support user preferences and seasonality in outfit suggestions.

• Developed and maintained a web application used by over 80 service centres worldwide.

• Contributed to both frontend and backend using React, Next.js, Nest.js, and PostgreSQL.

• Ensured cross-device compatibility and high responsiveness across all major browsers.

• Integrated scalable API services with authentication and database layers.

• Scaled the product from concept to 1M+ registrations, 60K DAU, and 20K MAU in 12 months.

• Managed a cross-functional team of 30+ (Dev, DevOps, GameDesign), introduced Agile workflows and CI/CD improvements.

• Increased average session time by 500% (10 mins → 1 hour) through game balance redesign and reward system revamp (ABC tests).

• Reduced churn rate by 15% using CJM-based onboarding, tracked with Hotjar and Amplitude.

• Localised the product into 8 languages, forming partnerships with local KOLs based on GEO analytics.

• Increased TVL from $5M to $28M (+460%) in 9 months by launching auto-compounding tools and APY incentives.

• Established partnerships with 5 major DeFi protocols to drive liquidity inflow.

• Boosted staking conversion by 45% by simplifying UX (7 steps → 3) and integrating 1-click staking.

• Used Dune Analytics and on-chain user behaviour data to inform roadmap decisions.

• Designed and developed a cross-chain bridge with support for 5 networks, optimising gas usage and reducing transaction costs by 30%.

• Integrated Aave and Uniswap V3, which resulted in 12,000 new users in one quarter.

• Reduced transaction processing time by 40% (12s → 7s) through smart contract refactoring and admin panel cache optimisation.

• Tracked results via Snowtrace and internal block metadata.

• Led the architecture and development of a Data Lakehouse platform for enterprise-scale data storage and analysis.
• Designed and launched a secure component analysis (SCA) system to scan, validate, and deliver safe software artifacts.
• Developed a custom ETL pipeline integrated with Greenplum and Airflow, enabling flexible, reliable data processing.
• Mentored three intern cohorts (Python + Go, DevOps, React) and initiated an internal R&D direction for the company.
• Participated in pre-sales and architectural evaluation of new projects, aligning tech decisions with business needs.

• Built a new development department from scratch, defined internal standards, and established the tech strategy.
• Oversaw in-house development that replaced vendor software with no loss of service quality.
• Successfully migrated two mission-critical systems from Oracle to PostgreSQL, significantly reducing licensing costs.
• Re-architected the company’s DWH, moving 70% of reporting to the new platform for improved flexibility and speed.

• Led architecture and development of several new products: payment aggregator, ticketing platform, parking system, and education accounting tools.
• Developed PL/SQL-based databases and documentation to support operations and compliance.
• Delivered two platforms reaching over 1M MAU.
• Boosted EBIT of the payment platform by 70% after implementing new services.

• Developed UI for a CRM system with integrated Stripe-based payment modules.
• Built a responsive, cross-platform interface.
• Integrated ContentStack CMS and ApolloGraphQL APIs for dynamic content and data fetching.
• Contributed to user account functionality and payment cabinet development in a 21-person team.

• Developed a thematic social platform with game-like functionality.
• Migrated codebase from Angular 8 to Angular 9.
• Integrated calendar API, improved performance (Lighthouse), and implemented SEO meta tags.
• Built responsive templates, multilingual support, and login via Google/Facebook.
• Deployed via Firebase Functions and structured local/session storage.

• Led architectural evolution of a large-scale product using microservices, aligning with business growth and scalability needs.
• Integrated and optimised component testing processes, enabling stable CI/CD and automated release cycles.
• Reduced infrastructure and development costs through architectural review and system optimisation.
• Introduced development and service metrics across the product to improve observability and accountability.
• Actively planned and scoped technical initiatives in collaboration with product owners and multiple teams, ensuring delivery of long-term goals.

• Designed and implemented payment processing flows, including custom gift card authorisation and capture logic.
• Developed auto-reversal services to handle payment anomalies.
• Contributed to admin-facing tools for managing refunds, returns, and layout improvements.
• Integrated Visa Checkout and Amazon Payment, expanding available payment methods.
• Enhanced system maintainability by contributing to REST APIs and backend infrastructure for advertising features.

• Developed detailed customer personas segmented by traffic channel to inform campaign strategy.
• Analysed user behaviour by building funnels and identifying anomalies across website events.
• Created monthly reports evaluating the performance of marketing campaigns, influencing strategic decisions.
• Built data marts in Google BigQuery and Greenplum to support reporting and ad-hoc requests.
• Designed reports from scratch in Looker Studio and Visiology, including migrating dashboards between BI platforms.
• Implemented a Telegram bot with Airflow scheduling and error logging to notify analysts about issues in delayed queries.

• Designed 5+ comprehensive Power BI dashboards for operational performance across departments, with 30+ daily users.
• Built core data marts in PostgreSQL focused on the "order" entity, forming the foundation for company-wide reporting.
• Automated daily, weekly, and monthly data exports for the operations team using Python scripts.
• Conducted regular ad-hoc analysis to support ongoing business needs and decision-making.

As an iOS developer in FG BCS, a leading investment company in the Russian market, I contributed to their app for private investors. We wanted to enable our users to top up the card balance with Apple Pay. There was no direct way to implement this flow, so we had to leverage a third-party service that created an order for the refill amount and transferred the payment to the client’s card using our API. The transaction processing and balance renewal on the screen were imitated.

As an iOS developer in FG BCS, a leading investment company in the Russian market, I contributed to their app for private investors. User was able to enter parameters of an account or a card they wanted to create (currency, type, name, etc) and after confirmation refill it from another card, account or with Apple Pay

Created and developed the application logic for bank product calculators (credits, savings, etc.) from data layer to presentation layer.

Our users received profit for selling their goods and services via our payment gateway in crypto. Many of them wanted to convert this profit into fiat and withdraw it to their bank account. Previously we used a 3rd party service that had a $30.000 threshold for crypto-to-fiat transactions, which prevented our new clients from giving us a try since they couldn’t trust us enough with the amount this high, and SMBs couldn’t afford keeping this amount frozen for so long. I researched and led the integration of alternative service with their API and our UI, that allowed us to lower this threshold to as little as $20 and onboard large clients that previously used our competitor.

The admin side of our service was outdated, lacked cohesiveness due to the initial desire to accelerate time to market, and there were constant complaints about things being hard to find there and simple tasks taking hours. Our business development managers also noticed that users often required personal help with onboarding during calls and wrote mailed similar questions to our support, which meant spending extra time and money. I led the redesign of 14 sections of the personal account, including the payment history, invoices/donations pages, mass payouts, balance top-up/withdrawal, wallets management, and various integrations.

Leveraged best practices to ensure seamless and quick sign-up and onboarding for new users.

Leveraged and combined several payment management systems in the mobile app to receive the desired payment flows: user-platform-user, user-platform-merchant, merchant-platform-user. Optimised the usage of 3rd party APIs and their sequence to receive the lowest possible fee.

Led the creation of an admin panel for a new social media startup from scratch. Took part in creating the project documentation, design, backend and frontend services with a product team. We created the means to manage content and its categories (with a 3rd party tool that helps moderation + the work environment for human moderators), users, payments (payment history, integrations dashboard), notifications, and editorial pages.

◦ Led a portfolio of more than 8 consulting projects and M&A transactions with a combined value of over $120 million ◦ Facilitated a strategic partnership between a Fortune 100 FMCG company and an Eastern European coffee producer, resulting in a doubling of the distribution chain ◦ Prepared exit strategy from a FinTech portfolio business into a leading financial ecosystem with a 3x value growth ◦ Directed a cross-functional team of 6 to develop and implement a 5-year fashion category strategy for a leading Eastern European marketplace to reach top positions in fashion e-commerce ◦ Established a new partnership pipeline with 5 non-competitors, resulting in 8 prospective leads over a 3-month period ◦ Provided ongoing support to the leadership team, including shareholders and the CEO, through comprehensive research and ad-hoc strategy development

◦ Developed and implemented a B2C service for high net worth individuals in the asset management sector from scratch by using product approach and successfully acquired 3 clients ◦ Developed the Russian market expansion strategy for a Swiss technology company and achieved 7% quarterly growth ◦ Created a digital product strategy that increased customer portfolio by 25% for Eastern European investment company

With extensive team discussions I employed three distinct methodologies, namely the Vader sentiment analysis, Bag of Words, and a Hugging Face BERT language model, for the purpose of tweet classification into two categories: "Disaster Tweets" and "Non-Disaster Tweets.". Using BERT-based language models, natural language processing (NLP), data analysis, TensorFlow, and Python programming, i made an in-depth evaluation to determine the optimal approach for the given task.

In collaboration with a fellow Data Scientist, we developed a powerful machine learning model. This model proved to be highly effective in analyzing medical data and accurately predicting whether a patient had a medical condition, achieving a remarkable accuracy rate of over 90%. Using Machine Learning and TensorFlow we facilitated the model's development. Data Analysis and Data Visualization techniques enabled us to gain valuable insights from the medical data, while Python played a pivotal role in implementing and fine-tuning the model, resulting in its impressive predictive accuracy.

Achieved an 80% accuracy rate in classifying celestial objects as Stars, Galaxies, or Quasi-Stellar Objects using a machine learning model that analyzes solar data. It was done through Data Collection, Data Preprocessing, Model Optimization, Iterative Improvement and Documentation.I Gathered a comprehensive dataset of solar data, including various features and attributes. I then Cleaned the dataset by handling missing values and outliers. Further Chosing appropriate machine learning algorithms for classification tasks and then Fine-tune hyperparameters of the model using techniques like grid search or random search to improve its performance. Lastly, Documented the entire process, including data sources, preprocessing steps, model architecture, and results, to facilitate reproducibility and collaboration.

Led the shift of analytical infrastructure, enhancing data resources and decreasing update errors by 90%. Increased identifiable traffic sources from 60% to 95%. Played a crucial role in reducing fraudulent responses from 28% to 1-4% by strengthening moderation and creating a fraud analytics direction. Contributed to a project reducing the cost of applications from groups by 70% and reducing vacancies in central Moscow from 20% to 6%.

Developed an integrated automated online and offline sales analytics system, utilizing various tools like GA, ExpertSender, Dynamic Call Tracking, Naumen, and 1C. Reduced BigQuery costs by over three times and improved the effectiveness of advertising campaigns by 8-10%.

• Gathered a team of BI Analysts and Developers from the scratch
• Created a data warehouse prototype using Qlik Sense and built a full commercial business reporting system from the ground
• Complete evaluation of the company's business reporting system, including financial reporting. Sources of financial losses and inefficiency have been identified and quantified.
• Provided recommendations to the operations office on sales enhancement activities and monitored implementation, which resulted in a significant revenue impact.
• Built demand forecasting model using XGBoost Regression

• Developed an automated Python Machine Learning model that provides post-evaluation for sales support programs and takes mutual cannibalization into account.
• Organized and led an ETL developers team that built Avon Russia's Cloud Data Warehouse, the only data source with consolidated and verified business data for operational analytics.
• With a team of ETL Developers Rebuilt Russian data instance of Global DWH in Snowflake with attached data flow interfaces.
• Initiated, developed, and implemented new Representative discount system and performance tracking dashboard in Tableau to promote productive sales behaviour
• Initiated and performed audit of existing business planning processes, suggested ways to improve demand forecasting accuracy
• Initiated and performed series of strategic analysis on pricing, average order, representatives’ costs and benefits to understand causes behind current business performance and highlighted possible business solutions to the top management and enabled more accurate sales predictions.

• Created a team and built a business insights and analytics function from the scratch
• Successful converted Marketing Research department to big scale Business Insights function, combining under one roof Marketing Research, Demand forecasting, Trade efficiency analysis & modeling, marketing planning and reporting.
• Successful development and set up of Demand Forecasting process across all 6 markets.
• Improvement of deployed Primary Research methodologies and internal data sources.
• Development and launch of new analytical reporting initiatives.

As part of my role at Time Research & Innovation, I Ied creation of World Health Aid (WHA), a multilingual real-time health advice platform designed to support disadvantaged individuals. WHA is a comprehensive solution that facilitates telemedicine services, bridging the gap between healthcare providers and underserved populations. It combines live patient and clinician-facing decision support systems, offering a range of paid and free services, including self-diagnosis screening through video and voice chat, multilingual instant translation, voice-command symptom search, real-time messaging, and secure record keeping.

In the corporate sector, text data flows swiftly, emanating from diverse sources in various formats, destined for multiple endpoints and actions. This continuous stream of data necessitates management through structuring, making it easy to interpret and analyse, and ultimately directing it to the appropriate user or applying the insights obtained. I utilised several large language models (LLMs) streamline and enhance this process:

• A question-answering LLM, enabling users to inquire about the content of each document within the flow.
• A summarization LLM, offering a concise overview of each document.
• An NLP model for sentiment analysis, determining the sentiment of each document.
• An NLP model that calculates an absolute similarity score among all documents.
• Clustering, which groups similar documents together and simultaneously identifies outlier documents.

In the dynamic realm of banking, ensuring effective anomaly detection is paramount. Utilising my expertise in state-of-the-art ML algorithms and advanced Deep Learning techniques, I tackled this imperative at Deutsche Bank, a globally recognized leader in financial services. The solution involved several critical stages:

• Identifying relevant data from various sources.
• Researching and testing optimal Machine Learning algorithms for tabular data analysis.
• Exploring and validating appropriate Deep Learning models for extracting insights from textual data (Natural Language Processing).
• Integrating the resulting models into the rigorous banking environment, complete with all necessary software engineering components: Python, Object-Oriented Programming, Algorithm Design and Complexity, Deployment, and more.

Market research is a domain teeming with diverse data types, including tabular, audio, and particularly text. This wealth of data can be utilised not only to enhance internal processes through automation but also to equip the end client with valuable new functionalities and use cases.

• The objective of this use case was to automate the labelling of open-ended questions from online market research surveys. Open-ended questions, which receive text responses, require further post-processing to extract useful information. I automated this process for one of the world's largest market research firms using a series of interconnected techniques:
• Utilising various forms of word embeddings, leveraging both Neural Network models and general techniques.
• Implementing clustering to group similar responses.
• Applying summarization and content extraction methods to distil and highlight key insights.

Over the last two years, I've been dedicated to promoting the integration of AI among scientists at the globally acclaimed marine research institution, Plymouth Marine Laboratory (PML). PML's core research program is focused on critical areas such as climate change, marine pollution, and sustainability. In a recent research study published in July 2023, I developed Machine Learning approaches to ascertain the primary factors influencing the phytoplankton carbon-to-chlorophyll ratio in the Atlantic Basin.These methods provided valuable insights into the environmental factors affecting phytoplankton carbon and chlorophyll levels in the corresponding figures.

With a 9-year tenure as a freelance Audio Technology Consultant, I crafted a wide array of technical solutions. These encompassed the development and implementation of new real-time audio effects for Arm Cortex M4 hardware, the creation of VST plugins with the JUCE framework, the design and integration of a real-time processing system using Beagle Bone Black and the Bela realtime platform, as well as contributions in musical sound synthesis and Machine Learning consultancy for an interactive musical project.

Dating apps, while popular, had users spending excessive hours daily, leading to notable frustration. Research from smalltalks.ai indicated that the average daily usage was a substantial 1.5 hours, with a staggering 70% of users experiencing anxiety post-use. To address this, I developed an application powered by the advanced GPT-4 artificial intelligence. This AI not only mimics human interaction but crafts initial messages based on individual profiles, engages in meaningful conversation, and establishes emotional connections using a meticulously curated pick-up database. Once a rapport is established, it proposes a real-life meeting and notifies the human user, thus streamlining the dating process while reducing user stress.

As CTO at PrimerAI, I developed an intelligent platform for personalized cosmetics suggestions using ML and AI, including OpenCV, GPT-4, and K-means clustering for skin tone detection and categorization. I employed a holistic approach in developing and implementing this web app, which analyzed users' skin tones using computer vision and unsupervised ML algorithms. My contributions covered all aspects of the project, including designing and building the frontend and backend of the web app, as well as administering Linux servers. This ensured a seamless, end-to-end product experience that catered to our users' needs.

GregTraider is a streamlined framework designed for automated online trading and trading strategy backtesting using historical data. It simplifies the process by allowing users to write a single strategy for both backtesting and online trading, eliminating the need to rewrite and adjust strategies for different platforms. The primary benefit of using Gregtraider is that you only need to write and adjust your trading strategy once for backtesting, and it is immediately ready for real trading. This saves you the tedious work of rewriting and testing your strategy multiple times.

At WildFoundry, I held the role of developer advocate, producing educational programming content through articles and YouTube videos while offering client support. A prominent GPT use case from my experience involved analyzing the apps that users ran on their devices. To overcome the challenge of manually examining tens of thousands of users, I developed a script that scraped user pages one by one and analyzed the content via the GPT API before storing the results in a pandas dataframe. This approach enabled an unbiased and comprehensive analysis of user applications, yielding accurate numbers and insights that outperformed traditional survey methods. The ensuing report allowed the identification of target customer groups and guided the company in shaping its customer relation strategy.

In this project, I focused on implementing small open-source models, such as Zephir-7b, on single-board computers like the Raspberry Pi. This setup enabled the creation of a robot controlled by an LLM. Working with such compact models presents numerous challenges; one significant issue is the reduced level of intelligence compared to larger models like GPT-4. To address this, the project required fine-tuning the open-source LLM to achieve satisfactory results.

As a computer vision expert, I was consulted to address the challenge of facial pore recognition for a beauty application. The problem involved accurately identifying and measuring individual pores on various skin types and tones. To tackle this, I designed an algorithm that could adaptively capture high-resolution images and extract relevant features. Leveraging machine learning, I trained a model on a comprehensive dataset that I helped gather, comprising different face images under various lighting conditions. The resulting algorithm could recognize and measure individual pores with high precision, providing personalized skincare recommendations to users, and thereby enhancing the app's user experience.

Over a three-year period, we worked closely with one of Russia's largest banks to successfully establish a robust AI monitoring system. This system operates with precision to swiftly address system failures by pinpointing their root causes and making well-informed decisions. The model expanded the scope of forecasted indicators to a total of 50,000, achieving an 85% reduction in false positive alerts. It also delivered a 20% improvement in precision compared to previous methodologies.

Led a team to develop an AI monitoring system that was capable of detecting anomalies in various business indicators. By closely working with one of Russia's largest banks over a span of 3 years, we designed and implemented an advanced backend system dedicated to the detection of anomalies across a wide spectrum of critical business metrics. This system significantly decreased the time and effort previously invested in manual incident detection and enhanced operational efficiency.

I coordinated a team in the development of two distinct automated incident detection models, which yielded substantial improvements in incident detection times, reducing them by an impressive 5 to 15 minutes. These models played a pivotal role in streamlining processes and optimizing the bank's operations.