Tasks

• Develop machine learning models to analyze client behavior, segment users, and predict preferences.
• Build a personalized recommendation engine for workout plans, services, and offers.
• Create a scheduling optimization model to align client availability with trainer resources.
• Integrate a large language model as the conversational core of client’s in-app virtual assistant.
• Fine-tune the LLM to reflect the company’s brand tone, gym-specific procedures, and customer support guidelines.
• Build an ETL pipeline to unify data from client’s mobile app, gym systems, and CRM/ERP into a centralized repository.
• Implement secure cloud storage for models, logs, and backups using AWS S3.

Challenges

Solutions

• Client data used for modeling was anonymized using salted one-way hashes; personally identifiable information was excluded from analytics workflows.
• Custom instructions, prompts, and reinforcement feedback were used to shape the assistant’s responses, ensuring brand consistency and accuracy in gym-related queries.
• Built a robust data pipeline to extract, clean, and centralize information from multiple systems, improving data reliability for ML models.
• Combined behavior prediction, segmentation, and recommendations into one agent that could adapt based on real-time user feedback and platform usage patterns.
• The mobile app was updated with a clear privacy policy and tools for users to view, export, or delete their personal data.

Outcomes

• Successfully deployed a secure and compliant AI assistant that recommends content, manages schedules, and engages clients in personalized conversations.
• Improved user satisfaction through more relevant recommendations and faster response times.
• Maintained data privacy and built trust with clients through transparent handling of sensitive information.
• Laid a scalable foundation for ongoing machine learning improvements and LLM updates.

Technologies Used

Tasks

• Build a machine learning model to forecast price movements across selected trading instruments.
• Develop an AI agent that interprets model outputs to inform automated decision-making processes.
• Integrate reporting functionality to generate and deliver performance metrics and KPIs.
• Implement an external trigger system (e.g., email and webhook) to automatically activate the agent and handle follow-up responses.

Challenges

Solutions

• Developed adaptive models with online learning capabilities to regularly retrain on new data and adjust to changing market trends.
• Integrated macroeconomic indicators and real-time event data into the model to improve responsiveness to market shifts.
• Used automated feature engineering (AutoFE) tools to rapidly test and refine a wide range of input variables.
• Worked closely with financial experts to vet model features and interpret market behavior accurately.
• Deployed advanced NLP pipelines to extract signals from news headlines, social sentiment, and regulatory updates.
• Hosted models on high-performance infrastructure, optimizing serving latency and throughput using FastAPI.

Outcomes

• Delivered an AI agent capable of handling real-time price predictions for multiple instruments.
• Enabled automatic activation via email or webhook, with the agent retrieving inputs, executing analysis, and sending back KPI reports.
• Reduced manual intervention in forecasting and reporting workflows.
• Created a scalable foundation for automated decision-support tools in financial operations.

Technologies Used

Tasks

Challenges

Solutions

• Used OCR to digitize and structure course content from PDFs and video transcripts based on predefined business logic.
• Implemented a vector database to index content semantically.
• Deployed a Retrieval-Augmented Generation (RAG) system using the o3-mini large language model to enable contextual responses to user queries.
• Built a simple, secure interface with Streamlit for learner access.
• Hosted all infrastructure in AWS Cloud for scalability and performance.
• Designed the solution as an MVP, with ongoing iterations based on user feedback.

Outcomes

• Reduced time learners spend searching for relevant materials by over 70%.
• Made course materials accessible via conversational queries, improving user interaction and satisfaction.
• Established a solid infrastructure foundation for future AI features.
• Continued platform development is supported by real-world usage insights and feedback from learners.

Technologies Used

Tasks

Challenges

Solutions

• Designed an OCR system tailored to edge devices, with lightweight, efficient algorithms.
• Focused on optimizing text detection and recognition to meet low-latency requirements.
• Implemented a dual-validation mechanism to ensure container serial number accuracy, significantly reducing the risk of incorrect readings.
• Developed a user-friendly interface for easy integration with the client’s existing container tracking system.

Outcomes

• Accelerated the logistics workflow, cutting time spent on container verification by 70%.
• Reduced manual entry errors, improving data integrity and operational efficiency.
• Enabled on-site, real-time container identification without reliance on cloud processing.
• Delivered a scalable solution that can be deployed across multiple facilities with minimal infrastructure changes

Technologies Used

Tasks

Challenges

Solutions

• Developed a Generative AI-powered platform with a ‘ChatGPT’ like interface.
• Enabled tailored responses to natural language queries of current IBC codes.
• Integrated key design parameters specified in ASCE 7-16 design standards.
• Incorporated IBC code tables and figures into tailored responses.
• Provided the ability to refine queries iteratively.
• Generated project-specific building code review reports.
• Ensured the platform is mobile and desktop ready.

Outcomes

• Transformed the way users access design requirements and insights.
• Improved efficiency in handling design compliance issues.
• Reduced project timeline and budget overruns caused by non-compliance.

Technologies Used

Tasks

• Build audience AI clusterization models based on behavior patterns
• Develop models for splitting audience by conversion probability
• Conduct in-depth audience analysis using data from various analytics platforms.

Challenges

Develop models that can optimally allocate budgets and estimate income generated from each marketing channel or advertisement. Additionally, conduct in-depth audience analysis using data obtained from various analytics platforms.

Solutions

• Developed and adapted a multiplicative marketing mix model
• Introduced two models for clustering the audience based on behavioral patterns and dividing the audience by probability of conversion.

Outcomes

• A multiplicative marketing mix model was successfully developed and implemented
• Two clustering models were introduced for more effective audience segmentation and targeting.

Technologies Used

Tasks

• Develop AI algorithms with dashboards and web portal development
• Implement and optimize CV algorithms for complex industrial environment cases
• Data preparation and annotation for industrial use cases.

Challenges

The project involved developing multiple algorithms and solutions. Additionally, implementing and integrating web portals and dashboards with the platform required extensive data preparation and annotation.

Solutions

• Designed and optimized CV algorithms for complex industrial scenarios
• Integrated real-time location systems (RTLS) and other sensors with CV outputs
• Developed dashboards and web portals for seamless usability.

Outcomes

• Achieved design, implementation, and optimization of CV algorithms for complex environments
• Successfully developed and optimized pipelines for performance improvement
• Delivered minimal involvement in day-to-day tasks, thanks to local leadership and efficient solutions.

Technologies Used

Tasks

• Provide expert services in pharmaceutical R&D for developing deep learning models for drug discovery, including predictions of organic compound solubility and interactions with proteins and DNA.

Challenges

Solutions

• Specialized service combining ML expertise with chemistry and drug discovery.
• Development and deployment of deep learning models to predict organic compound solubility and interactions with proteins/DNA.
• Utilization of extensive datasets and advanced algorithms for accurate insights.
• Seamless integration with existing R&D workflows to ensure minimal disruption.
• Continuous support and model updates for adapting to new data and research needs.

Outcomes

• Achieve a 15-20% reduction in research time and costs, accelerating the drug discovery process and improving the success rate of potential drug candidates.

Technologies Used

Tasks

• Develop an AI-driven ecosystem to enhance HPLC capabilities in pharmaceutical R&D, enabling peak prediction, automated integration, and compliance with regulatory standards and GMP.

Challenges

Solutions

• Develop an AI-powered ecosystem to enhance HPLC capabilities.
• Predict peak retention times and integrate peaks automatically per GMP and regulatory standards.
• Utilize ML models trained on historical chromatographic data for real-time predictions and integrations.
• Include a user-friendly interface for parameter monitoring and compliance.
• Seamlessly integrate with existing LIMS for streamlined data management and record-keeping.

Outcomes

• Achieve a 20-25% increase in operational efficiency, reducing manual workload and enhancing the accuracy of chromatographic analyses.

Technologies Used

Tasks

• Implement AI/ML solutions to forecast material consumption and sales, optimizing procurement, supply chain, and production quantities in the Pharma industry.

Challenges

Solutions

• Implement ARIMA and LSTM models for time series forecasting of material usage and sales.
• Integrate models with ERP and CRM systems for seamless data flow.
• Establish a robust data governance framework.
• Enable continuous model retraining to adapt to market changes.
• Enhance procurement planning and production schedules to reduce material waste.

Outcomes

• Achieve a 10-15% reduction in material waste, improving overall efficiency and reducing costs.

Technologies Used

Tasks

• Optimize data acquisition from multiple real-time sensors monitoring the aniline synthesis process.
• Develop a Deep Learning algorithm to detect reactor issues using multisensor data.

Challenges

Solutions

• A complex Neural Network model was developed to perform sensor data fusion and predict reactor hard stop.

Outcomes

• The developed Neural Network model introduced additional control capabilities and improved process visibility, resulting in a 40% reduction in reactor failures.

Technologies Used

Tasks

• Audit current IT architecture.
• Analyze reporting processes across departments.
• Recommend solutions for analytic function optimization.
• Gather requirements for reporting and analytics.
• Define master data systems for each category.
• Propose new IT architecture for system integration.
• Outline development plan for the new system.

Challenges

The main challenge was to develop a comprehensive strategy to automate data collection and reporting processes, integrate various data systems, and improve data-driven decision-making capabilities for Servier Ukraine.

Solutions

• Established unified reporting standards for all data categories.
• Created a single Master Data system per category.
• Enabled automated data exchange across systems to reduce manual tasks.
• Built an automated reporting system for recurring reports.
• Recommended MS Power BI as the primary reporting tool.

Outcomes

• Automated data collection and reporting.
• Enhanced data integration and minimized silos.
• Strengthened decision-making with predictive analytics.
• Reduced manual tasks, boosting data efficiency.
• Improved user experience with a modern interface.

Technologies Used

Tasks

• Develop a classification model to help customers understand their spending habits.
• Create an adjustable ML model to profile credit and debit transactions.
• Build a clustering ensemble to define regular and irregular transactions.

Challenges

The main challenge was to create a competitive advantage by utilizing AI to distinguish the financial app from others and to grow expertise in advanced analytics and machine learning.

Solutions

• An adjustable ML model that profiles customer transactions by merchant, purchase type, and income.
• A clustering ensemble that categorizes each user’s transactions as regular or irregular.

Outcomes

• The ML saving advisor became one of the most used features of the app.
• The first version of the model allowed accurate labeling of regular and irregular transactions in 80% of cases.

Technologies Used

Tasks

• Developed visualizations for client segments and profiles
• Predicted churn likelihood with machine learning models
• Set up data warehouse and database schema
• Automated data updates for Power BI dashboard
• Implemented alert system for data updates and exchanges
• Created tools for churn retention strategies
• Compiled detailed project documentation

Challenges

Building a robust data infrastructure and predictive models for effective churn retention.

Solutions

• Set up a data warehouse (likely AWS Redshift or Google BigQuery).
• Use scikit-learn models and gradient boosting for churn prediction.
• Automate data processes and implement retention strategies.

Outcomes

Improved client retention and data-driven decision-making through advanced analytics.

Technologies Used