Automated Data Engineering for Big Data Analytics

Sentenai's commercial solution is a data fusion and decision support system that automates the data engineering processes required to unify operational data under a single, optimized virtual context, ready to be used for interactive analysis by end-users to make real-time decisions based on analyses performed across thousands of data sources simultaneously. In complex operational environments with a large number of data sources, decision-makers are continuously forced to make the tradeoff between making a decision based on an analysis of all the data and making a fast decision based on only a small amount of the data. Sentenai takes an automated approach to address the data engineering challenge of combining thousands of disparate data sources to perform comprehensive, interactive analysis of data.   Sentenai’s proposed solution uses our proven AI backend commercial solution, Tempest, to automate the data preparation thereby accelerating data processing for analysis,  and ultimately dramatically driving down the cost in time, personnel, and data-to-decision. Combined with BAE Systems Inc.’s (“BAE”) All-Source Track and Identity Fuser (ATIF) automated fuser technology, Sentenai will work with Navy data sets to create a more efficient and effective analyst workforce dedicated to supporting high-level decision-making rather than routine data engineering tasks. As such, we believe further technology development under the subject SBIR topic will contribute to solving this mission need. BAE Systems’ All-Source Track and Identity Fuser (ATIF) is a TRL 9, state-of-the-art fusion engine that is or is in the process of being deployed on systems like the Navy E-2D and Air Force’s Network Centric Collaborative Targeting (NCCT). It performs tracking, correlation, and fusion of on-board sensors (e.g. radar, SIGINT, video tracks) and external sources (e.g. Link16, Thresher) using traditional methods from statistics, estimation and graph theory. The addition of Tempest’s Machine Learning capabilities to pre-process for input data issues, post-process for false alarms and erroneous fusion decisions, and feedback to modify parameters has the potential to significantly improve ATIF’s performance.  Sentenai and BAE Systems will develop a Proof of Concept (PoC) demo that runs ATIF with and without Tempest to showcase the potential for improvements by augmenting an existing fusion engine with advanced Machine Learning capabilities.