Technology Insertion Strategy and Architecture Development

GITI helps its customers to identify relevant technology from across the Intelligence Community, from commercial vendors, and from the research community. It is capable of developing conceptual visualizations to help explore the analytical value of specific technical capabilities and to help explore alternative ways to perform analysis. GITI also develops strategies and architectures for the exploitation of advanced technologies.

We help our customers to introduce more rigor into the analysis process.  We find, test, integrate, and deploy the tools and technologies that simplify the analytical process. When no tool is available, we find, develop, and integrate cutting-edge research from across academia, the national labs, industry, and government.  GITI’s traditional relationship with the research community is especially helpful in this challenging area.  GITI performs rigorous scientific evaluation of a particular tool or algorithm to identify strengths and weaknesses.  This gives GITI the opportunity to assemble the best combination of tools for our customer. 

GITI uses the integration of these diverse technologies to empower the analyst to propose a theory, then rely on our technology to search and reason with the available data to support or refute their theory. GITI applies best practices of Service Oriented Architecture augmented with Intelligent Software Agents to integrate diverse and independently developed analytical tools. We have successfully deployed an ensemble-architecture where we integrate competing and complementary products into an integrated system that significantly out-performs any one of the products. We are providing unique and powerful graphical tools that enable end-users to create ad-hoc analytical workflow that automate complex and time consuming tasks.

-back to top-

Rapid Prototyping and COTS Integration Using Agile Methodology and Spiral Development

Global InfoTek, Inc. has developed and employs an agile, experiment-driven process for rapid prototyping and integration. This process typically results in an initial capability within 90 days, and has involved development cycles as short as a week. The process allows us to focus on enhancing mission effectiveness and benefit to users by continuous refinement and early deployment of operational capabilities.

GITI is able to support rapid deployment of capabilities because of our broad, hands-on experience with relevant technologies, and our exploitation of proven processes and toolkits for integration.

GITI has a strong record of conducting cutting-edge, relevant R&D for DARPA, IARPA, and other research organizations. It is capable of developing solutions to unique operational challenges, and to rapidly transition mature research to operational settings.

-back to top-

Semantic Technologies for Improved Decision-making

GITI offers semantic technologies to enable our customers to derive more explicit meaning from their massive amounts of disjointed data. Our approach has resulted in better ways for enterprise software applications and people to understand, share and reason with large amounts of data for improved decision-making. Our solutions have enabled our customers to exploit their data in new ways, gradually adjusting the manner in which they are perceived, exploring new relationships and helping to make sense of non-obvious interconnected information. This encourages further linking and reuse to incrementally grow the available pool of data that can be shared.

Global InfoTek, Inc. helps our customers reduce the risk of, and accelerate the deployment of semantic technologies, through its extensive hands-on experience in developing and deploying semantic technologies. GITI has developed ontologies for the intelligence community customers to support information sharing and integration.

We have built semantic mediation solutions across multiple data sources originating across the intelligence and law enforcement communities. We have developed semi-automated tools to build semantic mediators for graph-structured data using a common ontology, substantially reducing the time and cost to integrate tools using such data.

We provide expertise with real-life deployment of semantic technologies that use World Wide Web Consortium (W3C) standards like XML, RDF (Resource Description Framework) and OWL (Web Ontology Language) that serve as a foundation for our approach.

GITI has combined agent-service, semantic graph, entity resolution, and other technologies (e.g., text extraction, machine learning, graph analytics, visualization) to provide a solution to acquire and fuse large volumes of data. 
GITI’s experience includes: building ontologies using Cyc, OWL, and DAML; building semantic mediators for military domains, exploiting “folksonomies” to generate semantic metadata; and assessing and building highly scalable RDF triple stores. This extensive experience allows our customers to exploit cutting edge semantic technologies to increase their mission effectiveness, while reducing the risks of adapting new technologies.

-back to top-

Knowledge Discovery and Management

GITI offers its clients significant experience in applying algorithms and tools to address the end-to-end knowledge discovery process where we extract knowledge from large volumes of unstructured and structured data. We bring to bear our intimate understanding of the characteristics of intelligence data, the COTS and GOTS knowledge discovery and data mining tools, techniques, and processes needed to ingest and analyze diverse sets of data for our customer's unique problems.

Specifically, we exploit intelligent pattern discovery and data mining algorithms to identify relevant patterns and develop models of the dynamic and adaptive environment , while also exploiting prior knowledge.

We bring to bear prior work and capabilities in pattern discovery, data mining, machine learning, agent-facilitated ad hoc team formation and support, user modeling, and exploitation of prior and tacit knowledge. We leverage on our significant experience and ongoing R&D activities in support of DARPA and IARPA to jumpstart projects.

-back to top-

Seamless Integration of Analytical Tools with Provenance Tracking

GITI uses agent and service oriented architecture to integrate diverse analytical tools at a significantly lower cost and effort where the traditional approach is based on creating customized one-to-one integration solutions. We have successfully created an analytic workflow management system that the analysts call the "Universal Connector."  The Universal Connector reduces the large amounts of time currently required to stage and process all-source intelligence data, then use different applications for discovery and visualization of this data.  Before our Universal Connector, only developers could run or alter an automated analytic workflow.  Now, analysts can run, schedule, and alter this analytic workflow without developer involvement.  We have integrated multiple analytic flows that ingest data and make it ready for routing to multiple analysis tools (e.g., Initiate, Palantir, NetOwl, Centrifuge, Spotfire).  Analysts can reroute these results for comparison with watch-lists using the combination of tools that are appropriate to a specific problem. The analyst is able to manage the analytic process without being exposed to the underlying technical details.

The Universal Connector infrastructure records data provenance throughout any ISL/CHAIN-enabled analytic workflow.  Provenance tracking is started immediately upon data receipt.  Provenance is recorded in detailed terms, such as data source, data record, unique personal identifiers (e.g., passport numbers, drivers licenses numbers), and the relationships between identifiers common to multiple records.  After receipt, the storage and analysis phases use this detailed provenance data to track the movement of data between agents in the analytic workflow.  Analysts do not have to consider provenance until they are ready because our ISL agent-services make legacy components “provenance aware” – they automatically register provenance data as they move data through the analytic workflow.  When the analyst has a provenance issue they need to investigate, the provenance meta-data our agents have registered is accessible through a semantic graphic query interface that can be used to explore complex provenance relationships.

-back to top-