ILTACON 2018 Panel: AutoClassification is Real; Don’t let Perfect be the Enemy of Good

Records & InfoGov professionals have long anticipated the ability to AutoClassify records and information. Vast information volumes make it unrealistic and inconsistent to rely on users to classify content – it’s time to let Machine Learning take over.


This session covers real-world case studies of successful AutoClassification implementations, including establishing effort scope, key decisions and considerations, and developing workflow logic and business rules.


Learn from your peers what’s possible now with current technology and what’s coming on the horizon.




The Importance of Data Loss Prevention

“DLP” (Data Loss Prevention, also referred to as Data Leak Prevention), is a term referring to the use of technology to protect confidential data from being shared with unauthorized parties. DLP systems monitor data in use, at rest, and in motion, seeking to prevent data breaches in real time. DLP technology relies on algorithms that determine which data transfers to block.

Why is DLP important? The value of information cannot be understated, and the risks associated with the exposure of sensitive data gives rise to the need for improvement in DLP practices. Due to growing concerns regarding issues such as: corporate espionage; cybersecurity data breaches; and changes in data privacy obligations, DLP technologies are a required business component for corporations and law firms. DLP protocols ensure that the flow of information within and outside of a corporate enterprise, or a law firm, follows an established path.

An element of DLP workflows requires interactions with other systems, such as a DMS (Document Management System), and/or a CMS (Content Management System). DLP protection relies on some other intelligence about the data in order to determine permissible transfers of information. DMS protocols established to determine where the files reside are often used to share information with the DLP technology, however, this does not generally address the specific content within each file. CMS technology can be used to identify content within files that require the DLP to block the transfer of that information. Determining “What” the file’s contents are is the first step in crafting a plan to protect such material. Files might require varying levels of security based on its content. Transfer of certain information might never be permissible outside the corporation by the DLP, while other transfers might be permissible but only under some limited set of circumstances.

Many corporations and law firms rely on their users to create and assign “tags” to newly created files. The level of sensitivity of a information may rely upon user classification to help determine the level of sensitivity of the individual file, or portions thereof. However, manual user generated classification is often inaccurate, and not an efficient process since it reduces employees productivity. Using technology to auto-classify documents help improve the DLP technology’s performance. Auto-classification is programmable to identify certain terms or phrases within a document that would trigger the DLP protections.

One of the shortcomings of DLP protocols is that there are a high number of “false-positive” incidents. These false-positive occurrences require I.T. involvement, and also delay transmission of information which can frustrate employees, thereby reducing productivity. By having files classified properly in a DMS, and using proper auto-classification technology to organize information, the frequency of false-positives that trigger the DLP protections is substantially reduced.

DLP systems are only effective if they have accurate knowledge about the data it’s trying to protect. Through the use of effective information governance and file classification practices, the performance of a DLP system can be dramatically improved. In addition, technology can be customized to add further enhancement to DLP practices. Certain files might be permissible for transfer if specific material is redacted. Technology can be employed to locate and auto-redact sensitive content. Through the use of automated redaction, DLP systems will permit transfers of data, while still preventing unauthorized sharing of specific sensitive information. Auto-redaction technology can reduce the burden that DLP systems impose on I.T. personnel by reducing false-positives, and also increase business productivity by allowing transfers of content that employees are authorized to share.

Valora’s unique proprietary technology, “PowerHouse”, serves to fill many of the needs that DLP systems require to function effectively. PowerHouse identifies the content of the data, and provides intelligence about each individual file or point of data. Valora’s technology is a “Rules-Based” system that can be custom configured to program the specific types of information that a corporation or law firm deems to require DLP protection. The Rules within PowerHouse include algorithms and elements of pattern matching recognition, and are used to auto-classify information, assigning categorization tags to files. The information classified by PowerHouse can be integrated into any DMS, and can also work in conjunction with other CMS software.

Valora’s PowerHouse not only works at the file level, but also at the content level within each document. Hence, the DLP can rely upon the auto-classification provided by PowerHouse to determine what information requires extra levels of protection. Using PowerHouse increases the efficiency of not only the DLP, but also enhances the performance of any DMS. In addition, PowerHouse increases business productivity by increasing the efficiency of data transfers. PowerHouse reduces the amount of false-positive incidents attributable to the DLP. In addition, Valora’s technology not only seamlessly enables the user to determine “What” the information in their possession is, but also helps enforce “How” that content needs to be handled by the DLP.

Moreover, PowerHouse enables the DLP to determine which individual users might have access to view transferred data, by enforcing established security permission levels. Hence, a DLP might permit internal transfer of information between certain individuals, but restrict others from having access to files, or specific portions of any document.

Since most DLP technology does not have the ability to determine what the contents of a file are, relying on PowerHouse to serve this function is an effective automated solution. In addition, PowerHouse is able to classify both structured and unstructured data. The classification provided by PowerHouse remains with each point of data, while it is at rest, in use or in motion.

Should you wish to learn more about Valora Technologies, and our proprietary solutions such as PowerHouse, please feel free to register at Valora’s website for our free resource information.

Guest Blogger: Joe Bartolo, J.D.

Follow Joe on Twitter: @joseph_bartolo and connect with him on LinkedIn

e-Discovery Journal – March 30, 2017 – DTI Invests in Valora Autoclassification

By Greg Buckles

It seems that I was not the only one interested in how Valora’s PowerHouse could address the sprawling corporate digital landfills. DTI has made a minority investment in Valora to provide the functionality to their Information Governance clients. In the M&A  world, a minority investment from what is essentially a giant channel partner is ‘neither fish nor fowl’to use a 17th Century idiom for something that is not easily categorized. We understand acquisitions, but Valora remains an independent organization from DTI and maintains its woman-owned status. The real question is how this will affect the nascent partner channel that Valora had just started to cultivate. Will DTI competitors be willing to use PowerHouse when they imagine that DTI can undercut them on the license fees? Will Valora essentially become a captive technology that is resold exclusively as a DTI service? This is essentially what happened to Patterns when FTI acquired Attenex. I don’t think that DTI’s investment comes at the price of Valora’s independent status, but I do think that the Valora team will have to work a bit harder to reassure potential channel partners or non-DTI customers of that fact. This investment confirms my new webinar slide (below) that shows DTI having the largest number of acquisitions/investments in the eDiscovery market. Early M&A Impact poll and interview results are starting to paint a picture of the concerns facing consumers when their provider is acquired or takes a large investment that changes their Go To Market strategy. So take my poll to see the results and join the ILTA webinar by Duane Lites and myself on April 12th to hear our take on how the eDiscovery market is consolidating and how you can mitigate the risks.

Stay skeptical my friends!



About Author

Greg Buckles wants your feedback, questions or project inquiries at He solves problems and creates eDiscovery solutions for enterprise and law firm clients. His active research topics include analytics, mobile device discovery, the discovery impact of the cloud, Microsoft’s Office 365/2013 eDiscovery Center and multi-matter discovery. Recent consulting engagements include managing preservation during enterprise migrations, legacy tape eliminations, retention enablement and many more.

Greg’s blog perspectives are personal opinions and should not be interpreted as a professional judgment. Greg is no longer a journalists and all perspectives are based on best public information. Blog content is neither approved nor reviewed by any providers prior to being posted. 


View Full Article Here

Press Release – March 29, 2017 – DTI Makes Strategic Investment In Valora Technologies


Global eDiscovery and Legal Services Market Leader Invests in AutoClassification Information Governance Pioneer

Atlanta, GA and Bedford, MA – March 29, 2017 – DTI, a global legal process outsourcing (LPO) company providing eDiscovery, management services, litigation support, and court reporting, and Valora Technologies, Inc., the leading innovator in AutoClassification, Predictive Analytics and Document Data Mining Technologies for Information Governance, eDiscovery, and Records Management, today announced that DTI has successfully completed a strategic, minority investment in Valora Technologies, Inc.  The investment marks the beginning of the coming-of-age of AutoClassification and the clear commitment of DTI to invest in leading-edge Information Governance solutions. Learn more.

March 29, 17|Categories: News|Tags: , , , |

LegalTech News – March 29, 2017 – DTI Looks for Next Step E-Discovery, Info Gov with Minority Investment in Valora

The investment allows Valora to focus on technology development over services, while DTI’s clients get access to data classification technology.

By Ian Lopez

E-discovery is a buyer’s market, defined by large companies like LDiscovery and OpenText buying up industry players to enhance their own service delivery. But the biggest purchase monetarily in 2016 was made by DTI, which bought Epiq in a deal valued at $1 billion that merged the two companies.

The company’s investment aspirations, though, did not stop there. Today, DTI announced a minority investment in Valora, a Bedford, Massachusetts-based technology and services company focusing on a variety of legal document management tasks.

As part of investment round in which Valora accrued $1.75 million in equity funding, the investment allows DTI and Epiq clients access to Valora technology for automated data classification and data mining. For Valora, the investment allows the company to shift resources toward technology development, part of an existing plan to focus more on technology than services. Neither company would comment on the financial specifics of the investment.

Valora will remain an independent company, and it had a relationship with DTI prior to the investment, auto-coding documents for in-house company and e-discovery projects, Kevin Jacobs, vice president of mergers and acquisitions at DTI, told Legaltech News. He noted that DTI was currently working on “active proposals” to provide Valora, and that the technology was already available to some clients.

“The purpose of the investment is to help advance the development in [Valora’s] categorization engine” to make it more product-focused, Jacobs added. Primarily a technology-enabled services company, “they kind of eat their own cooking if you will, and so we wanted to expand the product. We needed the product to be more market facing, and to streamline some of the processes to make sure it’s super effective for us to provide the services for our clients.”

In terms of DTI’s investment strategy, the company wants the technology to make the company “part of the day to day process of data management and compliance,” for clients, Jacobs said. He added that the company is “certainly looking” at other technologies but not discussing them at this time.

Valora’s “auto-classification” technology automates different tasks on the legal workflow, applicable to e-discovery, records management, litigation and information governance, for which it was chosen as a finalist for different categories at the annual InfoGovCon event in 2015 and 2016. Prior to the investment, the company was primarily “provisioning services on [its] technology core,” Valora CEO Sandra Serkes told LTN.

“What’s starting to happen is there’s more and more demand for the technology itself,” she said.

Among Valora’s services are document coding, review, intake and visualization; analytics; and hosting. DTI’s services include litigation support, court reporting, and managed services.

View Full Article Here


e-Discovery Journal – March 1, 2017 – eDJ Brief: Valora Tech

By Greg Buckles

It always fascinates me to see service companies dare to transform themselves into technology sales. So many of the earliest eDiscovery companies started in firms or scan shops who hired developers to meet client’s requirements when there was no acceptable Off-The-Shelf (OTS) software available. I was one of those corporate clients writing functional product requirement documents for my providers when regulators made the first demands for native email productions. Doing a briefing with Sandra Serkes about the evolution of Valora Tech was a trip down memory lane. Valora was founded as a typical service provider with an emphasis on records management. They developed custom automated document data mining, categorization and analytics software to meet client demands. The trick for any small custom development shop is to rise above the individual client requirements to create a solution platform that meets the pain points of the broader market. In Valora’s case, they went from automatic BIB coding to create their PowerHouse data mining platform to process and categorize live enterprise data. This is a lofty goal and I am the last person to hype something that I have not seen function in a client’s environment. Too many technology giants have overpromised and underdelivered on enterprise automated categorization over the last decade. So take my briefing notes with my usual ‘trust but verify’ advice.

Greg’s Take-Aways:

PowerHouse Platform

  • Process in place and extraction of 600+ metadata fields to create a rich/enhanced metadata database
  • On-premise or cloud implementation with remote connection without client side processing
  • Centralized management of data siloes – Exchange, archives, scanned paper, file shares, desktops, OneDrive, Boxx, DropBox, etc.
  • Recent partnerships with Alfresco ECM and iManage


  • Hosted visualization of enterprise and matter data collections
  • Taking PowerHouse metadata and making it multifaceted presentation
  • User window into PowerHouse database
  • Ability to dive into subsets

So why has eDiscovery driven much of the recent advancement in Artificial Intelligence, analytics and Predictive Coding (categorization)? Because we have real world problems with significant budgets and tight time frames driving our innovation. In my opinion, this is why many successful start-ups wither and die once acquired by global technology giants. They have no appetite for the blend of service-technology required to customers living in the trenches, so they lose the innovation edge. Always refreshing to talk shop with another eDiscovery Veteran of the Psychic Wars.

About Author

Greg Buckles wants your feedback, questions or project inquiries at He solves problems and creates eDiscovery solutions for enterprise and law firm clients. His active research topics include analytics, mobile device discovery, the discovery impact of the cloud, Microsoft’s Office 365/2013 eDiscovery Center and multi-matter discovery. Recent consulting engagements include managing preservation during enterprise migrations, legacy tape eliminations, retention enablement and many more.

Greg’s blog perspectives are personal opinions and should not be interpreted as a professional judgment. Greg is no longer a journalists and all perspectives are based on best public information. Blog content is neither approved nor reviewed by any providers prior to being posted. 

View Full Article Here

March 1, 17|Categories: News|Tags: , , , , |

Perspectives on Predictive Coding

Perspectives on Predictive Coding and Other Advanced Search Methods for the Legal Practitioner

ISBN: 978-1-63425-657-5
Product Code: 5310451
2016, 656 pages, 7 x 10

Reliance on manual and keyword search methods alone is increasingly seen as inadequate for searching large volumes of information. There are concerns about the accuracy and efficiency of these methods, especially as compared with more advanced search techniques. This book provides a set of perspectives on predictive coding and other advanced search techniques, as they are used today by lawyers in pursuit of e-discovery, in investigations, and in other legal contexts, such as information governance.

There is something in this book for everyone—novices, seasoned e-discovery practitioners, litigators, business lawyers, and technologists. It is meant to appeal both to practitioners who are seeking basic knowledge of what predictive coding and other advanced search methods are all about, as well as to those members of the legal community who are “inside the bubble” of e-discovery already and wish to gain further insight into the latest thinking on advanced search techniques from leading lawyers, judges, and information scientists. The book may also be read by lawyers who do not consider themselves litigators or e-discovery practitioners, but who wish to apply a knowledge of smart analytics in other legal contexts.

Valora CEO Sandra Serkes wrote Chapter 17, entitled  The Larger Picture: Moving Beyond Predictive Coding for Document Productions to Predictive Analytics for Information Governance

Purchase Book Here

December 19, 16|Categories: News|Tags: , , |

VALORA WINS TWO KEY INFORMATION GOVERNANCE AWARDS / Finalist IG Service Provider of the Year & Finalist Best IG Pitch 2016

Bedford, MA –  October 26, 2015 – Valora Technologies, Inc., a leader in predictive analytics for eDiscovery, Records and Information Governance, today announced it was named 2016 Finalist for Both Information Governance Service Provider of the Year & Best Information Governance Pitch of the Year at the Information Governance Conference in Providence, RI earlier this month.   As Best IG Service Provider, Valora received over 250 votes in online polling for its AutoClassification technology solutions for Records Management and Information Governance.  For Best Pitch, Valora received high praise from onsite conference attendees who voted live for their favorite suppliers who demonstrated a combination of technical acumen combined with industry leadership and enthusiasm in the marketplace.  [Read More…]

Load More Posts