-Providing consulting services for machine learning and advance analytics. Deep learning topologies and modeling. Entity analytics with Linguistic name matching (NLP with Machine Learning and Markovian Statistical Analysis. Rule based evaluation natural language structures with Prolog programing). Link Analysis construction graphs based on NLP analysis. -Architect Microsoft Azure with Enterprise Cloud Security, Cloud Datawarehouse, ETL, Storage, Kubernetes services, VMs, Docker containers. Azure cost optimization, budgeting and cost monitoring. ML with Saturn Cloud, AWS. -Azure Purview (Data catalog) implementation. -MLOps – architecture and development custom solutions (Airflow, TFX) -Time series research with LSTM (Long short-term memory with deep learning topology) and MCMC for time-series stationarity change detection. Using NLP context change over period of time (context-temporal) and its statistical analysis as predict factor for performance of business entities. -HPC programming with Julia (1.5) in Linux environment. Architecture HPC processes. -FinOps analysis and optimization for cost saving for organization (AWS and Azure). Cloud projects assessment for financial management during development and O&M phases. -Research with reinforcement learning with automated planning for intelligence automation/robotics.

Product owner for Link Analysis software (Social Network Analysis). Product included visual presentation of graphs from graph database and applied mathematics method for advance analytics. Managed team of 5 developers. (USCIS, DoD proposal and prototype) Analysis Information Diffusion algorithms in Social Network. Comparative analysis information diffusion in online Social Networks vs Per-to-Per Social networks (USCIS, DoD proposal and prototype) Predictive analytics with Time series multivariate analysis: statistical VAR and RNN machine learning. Labor demand Forecasting in law enforcements. (Python, Keras, Facebook Prophet, Statsmodels) Design MapReduce process for computation un-structural records (textual) similarity with fast approximate approach. Multistage data manipulation with creation cascading RDD for purpose generator cross-tab organized table containing elements encountering statistic. Enhancing Dimension Independent Matrix Square using MapReduce algorithm for fast (35% faster than standard MapReduce task) approximation. Design benchmarking process and optimization process with analysis’s data shuffling pattern between nodes and memory allocation with nodes with number RDD created. Deep Learning – MLN, CNN, RNN design and research

Leading R&D efforts for government client with developer and data scientist team. Design web-based tool for exploration Social Network Analysis in context of Risk Analysis. Visualization Geospatial locations and Temporal events with Graph Database, GIS, OpenLayer.js. Research on algorithms on Information diffusion in online and peer-to-peer Social Networks. Design Reinforcement Neural Networks for operational planning with constraints resolution (Deep Learning Reinforcement Learning, Prolog) Design and implementation Spatial/Temporal centrality ranking with centrality trajectory estimation of life-events. Predictive Analytics with Time series multivariate analysis: statistical VAR and RNN machine learning. Labor demand Forecasting in law enforcements. (Python, Keras, Facebook Prophet, Statsmodels) Research on simulation methods for construct (Deep Learning Generative Adversarial Networks) synthetic life-event histories of virtual persons for simulation for purposes training, testing and validation risk-assessment software and machine learning detecting processes. Design Geo Addresses analytics and statistical models. PostGIS based spatial analysis and features modeling. Data engineering (Data Architecture and Design, pre-processing, filtering, statistical evaluation of data set, aggregation) Design MapReduce process for computation un-structural records (textual) similarity with fast approximate approach. Multistage data manipulation with creation cascading RDD for purpose generator cross-tab organized table containing elements encountering statistic. Enhancing MapReduce algorithm for fast approximation. Research and troubleshooting for client performance problems based on Hadoop, Spark on distributed environment. Environment and working tools: Java, Python (PyCharm, Jupyter), MATLAB, simulation tool AnyLogic, PostGIS, GIS OpenLayers. Hadoop ecosystem (AWS EMR, Hive, Pig, Spark), Docker and application deployment with DevOps. AI/ML: Modeling and Features enginee

Research, Design and Development new technologies of applying AI science and Machine Learning in government funded R&D projects. Perform analyses and retrieving traveling patterns high risk commodities and passengers. Architecture and design Machine Learning components in “Enhanced Radiological Nuclear Inspection and Evaluation” project for detection nuclear materials on border crossing points. Collaborative efforts of Laurence Livermore National Laboratory and Carnegie Mellon University Machine Learning group. Organization and configuration independent verification and validation (IV&V) and Quality Assurance (Q&A) processing for Machine learning algorithms. Design intelligence automatization. Architecture and development distributed Machine learning processing in grid environment. Modeling and Features definition for Supervised, Unsupervised (EM, hierarchical clustering) and Semi-Supervised Machine Learning classification techniques (Kernel based SVM, Random Forest and its variation algorithms). Metaheuristics algorithms for optimization scheduling of Machine Learning training. Creation correlation and role-based models for detection high risk passengers acting in ring (swarm intelligence modeling). Open source based national security tread analysis and modeling. (Social Networks Analysis, Public Information analysis and summarization). Creating Entity Analytics entity resolutions software, including Linguistic Names matching, Ethnicity and culture detection, Address proximity Resolution, GIS Geographic Information System. Research on following topics: Effect of anomalies of floating-point arithmetic with subnormal values on performance Machine Learning models. Ensemble Machine Training methods and measuring performance learning model on “archipelagoes” distribution of multi-classes data sets.

Resolution) system for Risk and Fraud Detection Management module. The product use as core component for account integrity, detection illegible for US immigration benefits seekers and detecting national treats and fraudulent individuals. (IBM Infosphere ISII) Linguistic Name Matching algorithms, Global Name Resolution and rule-based system for providing immediate and actionable information to help prevent fraud and abuse of USCIS. Providing Statistical, Data Mining, Predictive Modeling architecture solutions. Design symmetric parallel processing with multiple computational threads for acceleration massive data processing for ultra-fast entity resolution process. Design statistical sampling selection methodology with approval of head of USCIS statistic department.

Lead a cross functional team to deliver Business Intelligence solutions (SAP BI) for financial and FAA General Ledger system. Development new methodology for analytical estimation and forecasting. Developments methods for optimization massive financial data processing. Design, evaluation and development technical specification for enterprise scale text search and invert indexing method. IBM internal project Product Concept Visionary for “Innovation through Prediction Market”. IBM Biztech software initiative. Received IBM management award. Multithreading Java predictive market engine.

Providing Data architecture design for Company key strategic and industry innovative project. Design Database Architecture and data processing components – data model for Enterprise scale Transactional system. Managed Technical aspects of Business Intelligence projects including statistical, regression analysis and algorithms. Design and development statistics algorithms for identification anomalies in radio broadcast signals for purpose remote diagnostic by analysis pattern of radio signals. The solution permit to reduce necessity in hardware, reduce cost and achieve greater degree of accurateness of results.

Design and Development Data Warehouse Architecture for Enterprise Occupational Health Management system: Analyzing Data Sources, Defining projects hardware and software components, Project risk analysis. Design Data description and Data flow organization. Conceptual, Logical and Physical Data Modeling. Identification criterions for data transformation. Design and development generic and cost-effective companywide solution addressed data transformation needs of projects with limited budget.

Design of Business Intelligence Reporting Systems with Advanced Statistical and data mining Analyzing techniques, Visual Data Representation methods for analysis of labor/material cost of aircraft engines services. Development Architecture and Data Modeling, conceptual, logical and physical database design for complex Data Warehouse system. Multidimensional OLAP design. Design, development and control technical aspects of the project for analysis effectiveness and responsiveness GE Aircraft Engines Customer Service Center and Phone Call center for processing mission critical, technical and sales support cases. Finding solution for projects cost reducing procedures by utilizing pre-exited components and modules. (Oracle, Erwin Cognos)

Technical project manager for Budgets, Financial and HR analysis Business Intelligence Systems for Department of Treasure of State of Israel, Dead Sea Chemical Industry Corporation. (Oracle OLAP) Provided Business Intelligence solutions based on Oracle BI methodologies and OLAP tools, while participating in pre-and post sale operations. Design, development and technical management of “Comparative Market Analysis and Competitors Intelligence” project for Israeli Aircraft Industry Corporation; “Labor effectiveness and Cost Analysis” project for Israeli Military Industry Corporation: Architecture, Modeling, Design, and Rollouts