Services
(Referrals are available upon request)
Domain | Industry | Service | Problem solved | Technology | Project lead(s) |
Business intelligence | Distribution | Curate equivalence/substitute relationships in products | Data visibility | Python, Plotly, Networkx | Josue Aduna, Ștefan Bîldea |
Business intelligence | Entertainment | Dashboards, automated reports | Data visibility | Tableau, Power BI, Shiny R | Josue Aduna |
Business intelligence | Finance & Banking | Dashboards, automated reports | Data visibility | Tableau, Power BI, Shiny R | Josue Aduna, Vlad Drăgan |
Business intelligence | Mobility | Dashboards, automated reports | Data visibility | Tableau, Power BI, Shiny R | Josue Aduna |
Business intelligence | Web services | Dashboards, automated reports | Data visibility | Tableau, Power BI, Shiny R | Josue Aduna |
Computer Vision | Healthcare | Visual odometry, 3D object mapping | Tracking progress in a fitness program | DL, Computer Vision | Bogdan Budescu |
Data Science | Distribution/ eCommerce | Solving data exploatation problems | Several | Python, R | Josue Aduna, Ștefan Bîldea |
Decision automation | Chemical industry | Price optimization | Find optimal price (range) | Client segmentation, LASSO regression | Ștefan Bîldea |
Decision automation | Distribution | Price optimization | Find optimal price (range) | Client segmentation, LASSO regression | Ștefan Bîldea |
Decision automation | Distribution | Recommendation engine | Extend warehouse product list | LightFM | Josue Aduna, Stefan Bîldea |
Decision automation | Finance & Banking | Statistical simulation | Risk assessment for dependent assets | Copulas probability modelling | Josue Aduna |
Decision automation | HR | Price optimization | Find optimal price (range) | Client segmentation, LASSO regression | Ștefan Bîldea |
Decision automation | Retail | Recommendation engine | Cross-sell/Up-sell | LightFM | Ștefan Bîldea |
Decision automation | Retail | Statistical simulation | Safety stock, optimal stock | Monte Carlo simulations | Josue Aduna, Ștefan Bîldea |
Forecast | Healthcare | Time series forecasting | Schedule optimizer for doctor's appointments in a clinic | Classification models in R | Josue Aduna |
Forecast | Accounting | ML model for invoice closing probability | Estimation of end of the month cashflow. Detect potentially late invoices to automatically send payment reminders. | ML classifier | Ștefan Bîldea |
Forecast | Accounting | ML model to predict days late for invoice | Estimation of end of the month cashflow. Detect potentially late invoices to automatically send payment reminders | XGBoost | Ștefan Bîldea |
Forecast | Digital marketing | ML model for user intent estimation | User intent based on webtracking | Deep Learning | Bogdan Budescu |
Forecast | Distribution | Time series forecasting | Safety stock, optimal stock, optimized warehouse mix | STL, liniar programming | Josue Aduna, Ștefan Bîldea |
Forecast | Finance & Banking | ML model to match trade with offers | Mortgage backed securities - trade reports are at TBA level for 18 months after which full disclosure ensues. During this time, to be able to use this data in pricing, one has to match the trade with the specific chain of intraday offers. | Deep Learning - MLP | Ștefan Bîldea |
Forecast | Finance & Banking | Price forecasting | Mortgage backed securities - bond price forecasting | Deep Learning - MLP | Ștefan Bîldea |
Forecast | Finance & Banking | Time series forecasting | Cash flow, financial forecast | Time series forecasting models in R | Vlad Drăgan |
Forecast | Finance & Banking | Time series forecasting | Cash flow | XGBoost, Deep Learning - LSTM | Ștefan Bîldea |
Forecast | Finance & Banking | Time series forecasting | Mortgage backed securities - TBA price forecasting | Deep Learning - LSTM | Ștefan Bîldea |
Forecast | Finance & Banking | Time series forecasting | Profit maximization in e-trading | Deep Learning | Bogdan Budescu |
Forecast | HR | ML model for Churn/Attrition | Find profile and context for churn/attrition events | Ștefan Bîldea, Lucian Sasu | |
Forecast | Manufacturing | ML model to predict faulty products | Reduce number of calibration rounds in manufacturing | Ștefan Bîldea | |
Forecast | Retail | Time series forecasting | Forecast demand | SARIMA, STL | Josue Aduna, Ștefan Bîldea |
Forecast | Transportation | Modeling remaining useful like of car subsystems | Predictive/preventive maintenance | AWS, Pyspark, ML modeling | Ștefan Bîldea |
IoT | Generic | Custom platforms for several categories of IoT devices | HW platform, middleware | Raspberry Pi, Arduino, Linux | Vlad Popescu |
IoT | Generic | Regression and classification models for industrial datasets | Several | Lucian Sasu | |
Proces automation | Finance & Banking | ML model to extract data from dynamic semi-structured data in emails | Automatization of information extraction about mortgage backed security offers sent via email | ML classifier | Ștefan Bîldea |
Process automation | Generic | Sentiment analysis engine | Information extraction from user feedback/comments | Gabriel Pelmuș | |
Risk management | Finance & Banking | Risk modeling | Financial profiling | Value at Risk models in R | Josue Aduna |
Software Defined Radio (SDR) | Telecom, Research | SDR platforms | HW platform, middleware | Ettus Research, NI, Linux | Vlad Popescu |
Stefan has a wonderful background mixed between science and computer science, which gives him the distinct advantage of not only being able to come up with creative scientific solutions to business problems, but also allows him to quickly create code to implement his solutions and test them.
From business insights to predictive analytics, providing actionable insights based on ML/AI, BDM provided a wide range of skills, knowledge, and solutions that lead to added value to our projects.
We needed automated data collection process from unstructured emails.
While the existing tools already provided automation aid for data entry, BDM identified, prototyped, and implemented a solution based on machine learning models to collect key data points from emails.
Stefan has supported me in a data-scientist capacity on several customer value measurement engagements over the last 3 years. While Stefan's has top skills in building complex predictive models linking multiple data sources, what stands out to me most is Stefan's focus in linking the data insights to the business narrative.