Dmitry Kaplan

Dmitry Kaplan, Ph.D


Areas of Expertise

  • Engineering and Project Management
  • Predictive Analytics
  • Data Mining
  • Machine Learning / Artificial Intelligence
  • Fraud Detection / Loss Prevention

About Dmitry

As a professional consultant, Dmitry understands that delivering maximum value in the shortest time requires quick, but thorough understanding of customers’ past development history, refinement of current requirements, and an appreciation of their vision of future growth.  With close to three decades of experience and dozens of clients, Dmitry is equally comfortable in a multinational and multilingual corporation and an angel-stage start-up; firmly believing that sound engineering and strong IP stance are the necessary conditions for long-term success at all stages of a company’s growth.

Dmitry and his lovely wife live in South Bellevue and are empty-nesters with two grown daughters – whom they are very proud of.  In his spare time Dmitry is a professional photographer and spends many a spare moment behind his Nikons, or in front of Photoshop.

Experience Qualifications

Dmitry started his consulting vocation while still finishing his dissertation at the University of Washington.  While his degrees are nominally in Electrical Engineering, he has progressed in the diplomas from Analog Electronics to Mathematics to Artificial Intelligence.  It is this desire to study as many disparate subjects as time and sanity allow that led him to a career in consulting.

His two main areas of expertise are Medical Device development and what he loosely terms “needle-in-a-haystack” problems.  The former includes many years developing medical ultrasound diagnosis and treatment systems, intelligent portable defibrillator algorithms, epileptic seizure warning systems and many others.

The second category encompasses more fields that all share the need to identify relevant, pertinent and actionable features in an ocean of informational nonsense.  Examples include:

  • Subtle fraud detection

Dmitry led the algorithmic development of a cellular fraud detection systems that reduced subscription and cloning fraud in major markets by 95%.  A similar, visual inspection-based approach was used in the identification of rare but persistent fraud in insurance submissions by health-care providers.

  • Recommender systems

Dmitry was a part of small team of scientists that developed the underlying approach to a group recommender.  This addressed the previously unanswerable question of what movie a group of people would most enjoy (or what restaurant).

  • Market segmentation

As the Business Intelligence lead at a major travel company, Dmitry classified potential travelers according to the destinations and travel suppliers (hotels, cruise lines, etc) they were most likely to purchase, reducing promotional costs by half and sharpening the focus of campaigns.  The ROI increase was at least 100% and often as high as 20x.

  • Customer conversion prediction 

Dmitry refined the feature pruning and selection approaches (as well as the decision engine) that predict with a high degree or reliability whether a potential web-site visitor is likely to buy the product.  This reduces (by a factor of three) the number of potential leads and improves their quality by the same amount.

  • HIPAA and Sarbanes-Oxley compliance

Dmitry developed algorithms for real-time identification of potentially damaging information intentionally or accidentally leaving the company.  The resulting system, integrated with Outlook/Microsoft Exchange, identified and required co-signatures for the email to be sent out.

  • Evidence detection

As a contractor to a legal evidence processing corporation, Dmitry developed visualization techniques and prioritization strategies that halved the number of documents subject to human review.

All of these problems share the characteristic of being data rich, but information poor.  It is this hunt for relevance that excites him the most.

Dmitry is committed to elegance in engineering solutions.  As a manager, lead or an individual contributor, he delivers maintainable, well-documented and robust solutions that are meant to survive the “first contact” with customers and which are easily refactored as computing environments, hardware architectures and software platforms evolve.