We are different from other software companies and consulting services groups in this important way: we use our own products to operate and manage our business.
We have virtually unparalleled experience launching
two, billion-dollar industries:
- automated legal processing — litigation support and law office management, spanning hundreds of projects, innovating all technologies.
- high-speed, high-volume, critical document-to-digital pipeline — built and operated a million-plus page/day facility; hundreds of projects, including CapitalOne and US House of Representatives digital mail handling.
Perhaps the best-known, most sophisticated document capture and management system on the market today is Captiva InputAccel, offered now by EMC. We conceived, designed and sold
Release 1.0 of that product to a company which EMC later acquired (precisely for their enterprise content management technology). We are among the founding architects of
high-flow-rate content capture pipelines. We have designed, sold and taught systems and procedures for, among others, Fujitsu, Xerox, IBML and Kodak’s respective digital imaging
operations. We have built systems to deliver and manage content in dozens and dozens of ECM systems, both general platforms like Documentum, and industry specific like Concordance.
We have worked with hardware and software companies to build capture and OCR/ICR systems with a demonstrated throughput rate in excess of 1,000,000 pages/day.
We believe that we know the issues and solutions involved in unplanned, unstructured (“let me get all you have”) enterprise-level data capture as well as anyone in the world.
We also have solid Microsoft .Net and SQL experience, and the practical knowledge which comes from having executed successfully hundreds of contracts in our service bureau operation.
Following are brief descriptions of just a few illustrative projects we have done (performing 100% of the work from inception through final data delivery). These are in no particular
order, but do illustrate the range of experience, project sizes and content importance we have:
Department of Homeland Security
- Developed a custom system, designed for operation by persons with a range of physical and other disabilities, with the objective of converting hardcopy mail into digital mail items and delivering them not later than four hours after receipt. We also were responsible for training the operators to use the system.
- During the first year of operation, the staff achieved a perfect score in quality measurements: 1) No defects in scanning the mail, 2) No late or missed deliveries and 3) No improperly addressed or mis-directed deliveries.
IHS Global Insight
- Custom software development for automatic capture and processing of oil and gas production data provided by state offices of Wyoming, North Dakota, Colorado and Oklahoma.
- Custom software development for automatic capture and processing and cataloging of federal government contracting opportunities for military applications.
- Capture and indexing of oil and gas well log files for the states of Wyoming and Texas.
- Improved profitability of the document scanning and archive management group of IHS in Houston, TX. In one quarter, we redesigned the document processing and warehouse management processes, adding 4 points of profit to the bottom line.
- Consulted on the development of a mineral leased-lands database for Oklahoma.
State of Kansas
- Processed 20 years of UCC and other corporate filing documents to redact sensitive and personal information.
- Performed automatic redaction using advanced OCR/ICR and text analysis for identifying redactions.
- Used geospatial algorithms to validate filing information for appropriate redaction processing.
- Delivered a high-speed software application for manual review with statistical analysis of the results and document signing for change tracking.
SRDS
- Built an online analytics system for processing and normalizing very large data collections about magazine and newspaper advertising.
- Increased the output from 30 available reports per month, to 300 per day, allowing for a new line of business.
- Produced a mass media ownership tracking system with live, fuzzy logic metadata searching across secure private collections, public data sources and realtime websites.
- Created Microsoft Office add-ins for data validation and customized purification, using their existing processes, but with more advanced data management.
Capital One
- Received, captured and converted 50,000 letters/applications per day.
- 24-hour maximum turn time, 7 days/week, 365 days/year.
- Automatic forms reading and name/address harvesting.
- The very first, and one of the few vendors ever to qualify for outsourced work from Capital One.
- Regulated by banking laws, these letters and applications required specific handling for confidentiality and speed.
PayMyBills
- A classic “dot-com” startup – one of the few to succeed financially.
- 35,000 letters per day – automatic content detection in unstructured data, routing and capture.
- Designed and built a hosting facility to support more than 100,000 simultaneous users.
United States Navy
- Captured all documentation about aircraft carrier and submarine, construction and maintenance. Much of this material was sensitive and NOFORN.
- A million or more pages of every type, from business documents to engineering drawings, with random, unstructured content on most.
- Delivered a secure, web-based classification and indexing system used by hundreds of military and civilian personnel distributed in various offices and installations.
United States Army Mortuary Services
- Performed automatic recognition and content-filing of the folders along with all their voluminous content, including fragile handwritten pages and photographs, of the death records from our Armed Forces personnel from WWII through the Vietnam War.
- Several million pages (difficult documents due both to their physical condition as well as their obviously very emotional content).
- Probably our most sophisticated unstructured content capture project, involving extensive text processing using regular expressions, and data-driven algorithms to assist in creating an ordered record collection from “documents” that ranged from photos of dog tags to Presidential citations.
United States Supreme Court
- Captured and converted to an electronic records system, the documents from five separate locations in the landmark lawsuit involving the states of Colorado, Nebraska and Kansas.
- Difficult, original documents including delicate/large maps and other one-of-a-kind pages.
- Over a million pages captured on site, in the field and loaded remotely.
United States Securities and Exchange Commission
- Designed and wrote a sophisticated, web-based, content capture and filing system for the internal tracking of investigative and litigation-related materials at the SEC, including all their remote offices throughout the United States.
- The security required was intensive, as the information contained in this system related to active litigation in U.S. District Courts, as well as highly-sensitive data about SEC probes into the activities of public companies. Accidental disclosure could mean millions of dollars in losses on Wall Street.
United States House of Representatives
- Designed and delivered the imaging system which moved constituent correspondence from paper form to an all-electronic format for Congressmen so they would not come into contact with potentially contaminated documents.
- Constructed a contained, negative pressure clean room facility with stand alone air handling and real-time air monitoring systems. This 10,000 sq foot facility was built to handle a variety of hazards, and we developed an expertise in clean-up procedures and handling potentially contaminated materials.
- There were five vendors who proposed; all billion dollar companies except us. We won on innovation, speed-to-deliver and demonstrated performance.
United States Department of Health & Human Services
- Designed a complete document capture system and processed millions of records consisting of OPF (Office of Personnel Files) files, converting them to the U.S. Office of Personnel’s “eOPF” standard.
- Much unstructured content; extensive use of .Net text-processing technologies to find, extract, format and appropriately file many, many fields of data.
- The system we delivered significantly reduced both the manpower and cost of this previously very laborious task.
United States Department of Energy
- Went to Carlsbad, NM, set up on-site a complete scanning facility, captured, indexed and delivered 500,000 pages of nuclear waste-related materials in 3 weeks flat. Hired and trained local personnel who had no previous experience in content management technologies.
- Built on-site (from our code templates) and used a completely custom app to fit into their existing technology environment.
- Our performance there subsequently qualified us as a sole-source supplier to the DOE in Washington, DC, to process 25,000 medical records folders (1M+ pages) of citizens claiming injury from a nuclear source.
- Much earlier, we also automated all of the classified records from the former Atomic Energy Commission’s bomb tests on Bikini and Eniwetok Islands.
State of Wyoming
- This is a project done on behalf of IHS.
- We “recovered” an unhappy relationship with the State and installed systems which still operate today.
- Helped them redesign and review their entire infrastructure, including databases, security and networks.
- We taught the client many technologies which have facilitated: (a) daily turnaround of transactional documents; (b) time-sliced access to their systems to regulate web traffic; (c) cutting their data entry costs in half; (d) developing e-permitting technologies; (e) the use of improved spherical trigonometry calculation methods which reduced distance errors from meters to inches.
Avaya, Inc.
- Rescued a failed FileNet installation by designing and delivering a web-based system for managing contract processing that was distributed worldwide, throughout the Avaya enterprise.
- Processed incoming documents from around the world, from all over the Americas, to Europe and Africa, to the Middle East and Asia. Converted, indexed and loaded these documents into their FileNet system, hosted in Denver, from our facility in Virginia within hours of receipt.
Central Pension Fund
- Processed many millions of pages from 90,000+ individual microfiche cards. These contained the entire active and archive records of the Fund.
- To facilitate ongoing operations during our process, we indexed all microfiche cards onsite, and delivered a web-based tracking and management system that allowed Fund employees to request expedite processing on any individual record. Before our work, it took them 8-10 hours to retrieve a single record from the microfiche collection. During processing, we could retrieve, convert and deliver the electronic record in less than 30 minutes.
Bush Cheney ‘04
- Given our pioneering work in Digital Mailroom technology, the campaign contracted with us to handle all the incoming correspondence.
- Processed many millions of envelopes, with an incredibly varied content, from handwritten notes and crayon drawings to official invitations and correspondence from dignitaries.
- Sophisticated indexing and automatic classification to route the data to specific personnel or departments for further handling. A great deal of security was required to ensure the confidentiality of the materials.
- Delivered a highly customized web-based mail client, similar to Outlook Web Services or Gmail, specifically for handling, tracking and processing image-based mail items.
LandAmerica Title
- Automated the capture of millions of land title records for the largest company in that business.
- Built what was, at the time, the largest online commercial image database in the world (over 8Tb).
- Worked directly with Microsoft on scalability issues with SQL Server and the NTFS file system. Designed new schemes for working with the issues involved in very large file systems and high-speed processing, including backup and disaster recovery issues.
Johnson & Johnson
- Assignment involved processing 17 million “Life of the Corporation” documents. These were both sensitive and critical documents, as many were related to the testing and trials of medical procedures and equipment.
- Involved working with documents of every type, automating the capture of both structured and unstructured information.
- Demonstrated an “exceptional level of security” capabilities (their words) in the process, as audited by their worldwide director of security, a former executive at NSA. Most of the documents we handled were covered under CFR Part 11 and HIPAA.
Several hundred law firms and their corporate clients
- We have undertaken engagements from some of the largest law firms and corporations in America, and have automated the capture, categorization and content management on well over half a billion pages. Some notable examples include:
- AT&T – the “breakup case” – the largest lawsuit ever undertaken (tens of millions of pages).
- AT&T vs. Bell Atlantic (now Verizon), a highly technical case that was largely determined on the quality of the indexing and classification of the documents.
- The McDonnell Douglas DC-10 crash cases involving documents of every possible description (50+ million pages).
- Two death-penalty cases (obviously high-value content).
- Many, many cases involving the RTC (Resolution Trust Corporation) involving bank failures (rapid capture and assimilation involving sources who were uncooperative, and judges who were impatient).
- Ford Motor Company – automated their entire Office of General Counsel (millions of pages; unstructured content; one of only three companies out of over ten thousand who were not required to undergo a “re-compete” during Ford’s difficult economic times.
- These assignments involved capturing information from every kind of source (e.g., web, electronic, paper, film, video, etc.)
- We have delivered both structured and full-text metadata, along with abstracts, to virtually every kind of technical platform and retrieval software package).
This is a brief discussion of the applications we developed in the course of our work across various industries.
Kaladi
Kaladi was built to gather well-related data from various state websites, such as Wyoming, North Dakota and Oklahoma. The first version was built in a matter of weeks from ArtaVia components
to gather and parse data directly from rss feeds, web pages and TIFF images, using multiple threads to enhance speed. Subsequent versions migrated to a production application that
captures a combination of web page data, web services and images. Kaladi also includes email notification and graphical reporting for both summarized activities and business intelligence
mining, as well as identifying potential system issues. Every operation, both manual and automatic, is logged for auditing and reporting, and system access can be carefully
administered through a rigorous, secure login process.
Epigri
Epigri was written to support the IHS Haystack product, and was the primary source to capture data and files pertaining to U.S. Federal contracts and RFP’s/RFI’s. The application is
designed around a plug-in architecture, with a central program that handles logging and execution. These plug-ins, built from existing ArtaVia components, retrieve files from FTP
sites, extract and parse data from web pages, follow web site structures to find new pages, extract compressed data, and process ASCII and PDF data.
Latrigato
Latrigato is a suite of applications and components designed on the ArtaVia framework for capturing data from websites and other online electronic sources, and then evaluating the changes
against historical references. Essentially, these are the same tools used in other applications, but further encapsulated to simplify and streamline the process of retrieving web-based
information and performing custom business-logic on the results. The Latrigato “suite” also includes a Transactions application, along with the corresponding underlying classes, to analyze
data for changes. This is a generic capability to evaluate data elements and build reference tables to identify changes and transaction events.
Foris
Foris is a document scanning process management system, built on the service bureau model from the ArtaVia framework. It includes all the typical components for scanning, ICR/OCR
(with unique enhancements), image enhancement, and forms processing, but adds significant capabilities for tracking and reporting on process metrics, efficiency and project planning.
Further, it adds sophisticated tools for material tracking, indexing, document preparation and integration with dozens of content management systems.
Tai’sen
Taisen is an application for non-technical to use to configure and construct advanced, web-based applications in the ArtaVia framework, with extensive validation and extensibility
through data binding and dynamic reflection. This can include both basic, client-based validation using Ajax and JavaScript, as well as complex, .NET workflow functionality. All of
this is accomplished with an interface as simple to use as Microsoft Excel or Access, but with the power and scalability of enterprise systems.
Legatus
Legatus is a web-based application, built on Tai’sen technology and ArtaVia components which supports online filing of state forms, such as well completion reports and sundry notices.
The application can be expanded by the state agencies themselves, to support filing of other forms like well production and sales data, applications for permits to drill, and change of
operator notices. Legatus creates Adobe Acrobat PDF files with smart, 2D barcodes that encode form data and security features to ensure the authenticity and fidelity of the submitted
forms.
Redact
Redact is a product for either manual or automated redaction processing. It allows an operator to process a large file structure of images, with extensive logging and change tracking.
The automated processing can redact images based on pre-configured zones, or on OCR results. These OCR results are further refined by Transformagic’s TextPurgator process to enhance
the quality of the result, with geospatial awareness of certain address information, for more precise data redaction.
Cerno
Cerno is a product designed for quality control review and indexing of image files. Cerno displays images in a grid of 4, 6, or 9 images displayed simultaneously so that the operator
may evaluate the images in context. Cerno is used to process images scanned and imported from an external source, such as microfilm images scanned by a third party, and is particularly
effective in automatically detecting the boundaries of documents based upon blips, form ids, or other features which may not have been captured in the scanning process.
Names Standardization and Management
The usability of public data is impaired by lack of standardization of names and other identifying data. We built a names standardization and management tool for a well-established
data services company. This tool improves search results by allowing the administrator to flag different names as equivalent for the purpose of searching and reporting, without altering
the underlying source data.
Our generic skills cover a broad array of technologies, industry domain expertise, and management experience. Here are some of the many proficiencies we bring to the table:
- 1D and 2D Barcodes
- Adobe Acrobat PDF and image management
- Advanced text manipulation and Regular Expressions
- Astronomical calculations - Microsoft.NET classes
- Astronomy & spherical trigonometry - Microsoft.NET classes
- Astronomy sun, moon, stars - rising/setting, sighting, shadows- Microsoft.NET classes
- Box labels and inventorying
- Complex geospatial applications - Microsoft.NET classes
- Complex process management
- Complex UI controls
- Content management
- Data cleansing and standardization
- Data mash-ups
- Document imaging
- Forms recognition and forms design
- High-speed data coding and indexing forms
- iPhone/iPad web apps
- Microsoft Excel data management, extraction
- Microsoft Excel functions and add-ins
- Microsoft Office add-ins and custom filters
- Microsoft SQL Server management
- Microsoft.NET Workflow
- Names management (large collections; standardization)
- Navigation/GPS - Microsoft.NET classes
- Network and infrastructure design
- OCR/ICR technologies
- Paper-to-digital mail processing
- Project management, automated reports
- Reflection, Windows internals
- Scalable database design and integration
- Screen based video documentation
- Teleprompters
- Text and image redaction
- Voice recognition and text-to-speech apps
- Web harvesting of data
- Web http filtering (high-speed search of websites)
- Web video and audio broadcasting