Tuesday, August 8, 2017

Charles River Analytics Inc., developer of intelligent systems solutions, has announced a contract awarded by the Defense Advanced Research Project Agency (DARPA) to build a virtual data scientist assistant, named Eve, to aid in solving real-world analytical problems. As part of DARPA’s Data-Driven Discovery of Models, or D3M, program, Charles River will lead a team on the Eve effort that includes Oregon State University & Win-Vector, LLS. The four-year contract is valued over $2.8 million.

In the D3M program, DARPA seeks to simplify the complex process of building models by creating software to bridge the data science expertise gap. Virtual data scientist assistants like Eve aim to help with machine learning tasks, and alleviate the shortage of data scientists needed for data-driven solutions now and in the coming years.

“A lack of qualified data scientists is preventing organizations from gaining the full benefits from data analytics and machine learning,” explained Dr. Mukesh Dalal, Principal Scientist at Charles River and Principal Investigator on the effort. “Eve will allow non-data scientists, such as retail marketers, to get significant value out of their data sets and make high-quality business-critical predictions.”

In the Eve effort, Charles River is fusing data-and knowledge-driven approaches for building representation formalisms and developing artificial intelligence-based search methods for planning and optimizing complex pipelines, incorporating data analytics and machine learning operators. Eve’s capabilities will enable subject matter experts to solve complex analytical and learning problems, without requiring assistance from data scientists.

Eve complements Charles River research and development in the Model Analyst's Toolkit, a software tool to analyze and validate scientific theories and models using real-world data.

Learn more about our related efforts in decision aiding, machine learning, and automated planning, or contact us.


This material is based upon work supported by the Air Force Research Laboratory (AFRL) and Defense Advanced Research Projects Agency (DARPA) under Contract No. FA8750-17-C-0120. Any opinions, findings and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the United States Air Force or DARPA.

Distribution Statement "A" (Approved for Public Release, Distribution Unlimited).

Close