Can ‘Predictive Coding’ cope up with ‘Big Data?

Discovery has changed, and electronically stored information (ESI) was the facilitator. Though ediscovery matters are no longer the novel issues that they once were,” technology is constantly changing. According to Baseline, it was estimated that 90 percent of worlds data has been created in the last two years.  in 2009 there were 988 Exabyte of data in existence, an amount that would stretch from the Sun to Pluto and back in paper form. The problem for corporations is the storage of huge amounts of data – let alone worry about the ‘compliance monster’.

Perhaps, cloud computing is here to ease things out, yet companies are retaining more information than ever, and lawsuits sometimes require attorneys review millions and millions of documents. While Judiciary struggles to devise effective mechanism regarding proportionality rules, big data is growing even bigger – not to mention growing litigation industry. It seems manual review of documents is not an option anymore, as technology is rushing towards meeting the growing needs of document review.

The most important element overlooked is the fact that human eyeballs are still required to review such documents leading to defensibility of the case; after all, isn’t that the real objective?

Definitions of “predictive coding” vary, but a common form of predictive coding includes the following steps. First, the data is uploaded onto a vendor’s servers. Next, representative samples of the electronic documents are identified. These “seed sets” can be created by counsel familiar with the issues, by the predictive coding software, or both. Counsel then review the seed sets and code each document for responsiveness or other attributes, such as privilege or confidentiality. The predictive coding system analyzes this input and creates a new “training set” reflecting the system’s determinations of responsiveness. Counsel then “train” the computer by evaluating where their decisions differ from the computers and then making appropriate adjustments regarding how the computer will analyze future documents.

This process is repeated until the system’s output is deemed reliable. Reliability is determined by statistical methods that measure recall—the percentage of responsive documents in the entire data set that the computer has located—and precision—the percentage of documents within the computer’s output set that are actually responsive. (That is, “recall” tests the extent to which the predictive coding system misses responsive documents, while “precision” tests the extent to which the system is mixing irrelevant documents in with the production set.) The resulting output can be either produced as is or further refined by subsequent human review. Subsequently, attorneys review a much smaller set of documents. Predictive coding therefore effectively “alleviates the need to review whole masses of records in order to find the relevant few.” Most importantly, predictive coding is estimated to reduce ediscovery costs as much as 40% to 60% while maintaining search quality.

A statistic quoted in an IDC and EMC report says that the digital universe is doubling every two years, and will reach 40,000 Exabyte (40 trillion gigabytes) by 2020. The question is: Can predictive coding cope up with big data?

e-Discovery | cloud computing
New Jersey, USA | Lahore, PAK | Dubai, UAE
(855) – 833 – 7775 (703) – 646 – 3043

Please follow and like us:

Managed Review vs. Staffing Model

There are two standing industry models for outsourcing e-discovery document review projects – the managed review and the staffing model. While selecting a review solution – either staffing of managed, the solution should reflect an approach suffused in an understanding of applied best practices. Having said that, the following sets forth a minimal, standardized, framework which can and should be adapted to meet the needs of specific cases.

Project management

· Ensure a project plan is specifically designed and crafted to the specifications and requirements of counsel

· consistent with best practices

· Deliver a key set of documents or review protocol that govern the execution and project

· management of the review process along with workflow design

Team selection and training

· Staff personnel with expertise in specific area of laws relating to project. Develop specific job descriptions and define a detailed protocol for recruiting, testing, and selection

· Ensure conducive environment as well as conduct reference and background checks

· Preferably engage experienced personnel

· Ensure the review team receives comprehensive substantive and platform training


· Design processes, assignments and quality assurance steps specifically tailored to the project’s requirements

· Demonstrate compliance with key security and quality standards while maintaining acceptable pace

Quality control

· Adhere to six sigma principles

· Develop effective quality control processes to achieve key project goals

· Test first review work product using sampling method, targeted re-review, and validations searches

· Conduct statistics to ensure the highest quality end result

· Maintain auditing and track performance of reviewers


· Develop a formal schedule of communications with counsel adhering to laid down schedules

· Calibrate initial review results, seeking counsel’s guidance to confirm or correct results and to conform review protocol and training materials to insights gained


· Deliver regular, comprehensive reports to monitor progress and quality and to assist counsel in managing the review process

Productions and Privilege Logs

· Prepare privilege logs in accordance with specifications set by counsel

e-Discovery | cloud computing
New Jersey, USA | Lahore, PAK | Dubai, UAE
(855) – 833 – 7775
(703) – 646 – 3043


Please follow and like us:

E-Discovery: Document review value proposition you don’t want to miss!

The biggest cost driver in e-discovery is the document review part, where millions and millions of documents must be reviewed for potential relevancy and/or responsiveness. KPMG estimates that first level document review encompasses anywhere between 58% and 90% of total litigation costs. While predictive coding technologies have been somewhat successful in culling down most of the documents by utilizing an optimal combination of ‘recall’ and ‘precision’ values, human eyes are still required. Attorneys, senior paralegals power up review centers specifically designed for such projects only to find repetitive work day in day out – resulting in high turnover.

Typically, teams are rounded up and dismantled on an as-needed basis using attorneys, paralegals and law school graduates. This, in turn, has created a ‘day laborer’ mentality, as most projects are short term in nature, anywhere from a few days to months. Therefore, it is not uncommon for document review teams to start with 30 to 50 individuals, and during the course of the project to lose over half of the original members. With costs and inconsistencies in mind, high document reviewer turnover has not only affected the consistency of first level document review projects, but has led to an inefficient model of ‘training and re-training’, consequently resulting in escalation of costs.

Presently, though, due to influx of attorney workforce, pay rates have been seen to take a downward turn – a simple supply and demand situation.

The overall costs, however, remain high due to advent of ‘Big Data’. The question remains: How do you further reduce costs? There is good news!

Let me illustrate savings for a law firm or corporation in a simple hypothetical scenario:

For a 100 GB project (approx. 15,000 docs/GB – source, we are talking approx. 1.5 million docs @$40/hour or $.80 per document (assuming 50 docs reviewed per hour – source Total billing = $1.2 million. With offshore attorney rate @$20/hour or $.40 per document. Total billing = $600K

Savings for the law firm of corporation by 50%, that is, save $600K – flat!

One can only imagine the cost savings in much larger engagements. These savings provide additional value and act as supplement to the predictive coding technology – together they form a true win-win solution for clients.

e-Discovery | cloud computing

New Jersey, USA | Lahore, PAK | Dubai, UAE


(855) – 833 – 7775 (703) – 646 – 3043

Please follow and like us:

Difference between Attorney, Barrister, Lawyer, and Solicitor

Have you ever wondered? The difference between attorney, barrister, lawyer, and solicitor? How about advocate, counsel, counselor, or counselor-at-law?

Lawyer is a general term for a person who gives legal device and aid and who conducts suits in court. An attorney or, more correctly, an attorney-at-law, is a member of the legal profession who represents a client in court when pleading or defending a case. In the US, attorney applies to any lawyer. In the UK, those who practice law are divided into barristers, who represent clients in open court and may appear at the bar, and solicitors, who are permitted to conduct litigation in court but not to plead cases in open court. The barrister does not deal directly with clients but does so through a solicitor. The word attorney comes from French meaning ‘one appointed or constituted’ and the word’s original meaning is of a person acting for another as an agent or deputy. A solicitor would be the UK equivalent of the US attorney-at-law. Counsel usually refers to a body of legal advisers but also pertains to a single legal adviser and is a synonym for advocate, barrister, counselor, and counselor-at-law. As to the abbreviation ‘Esq.’ for ‘Esquire’ used by some lawyers … it has no precise significance in the United States except as sometimes applied to certain public officials, such as justices of the peace. For some reason, lawyers often add it to their surname in written address. However, it is a title that is specifically male with no female equivalent, so its use by lawyers should fade away.


About Claydesk:

Korporate Solutions, Inc. d/b/a ClayDesk is a leading provider of e-Discovery and cloud computing solutions worldwide with its registered office located in Lahore, Pakistan, and branch offices in Piscataway, NJ and Dubai, UAE. Our e-Discovery services cover the entire EDRM cycle including information governance and compliance. Having JD’s as part of our core team, our onshore and offshore document review capabilities are unmatched in the industry.

e-Discovery | cloud computing

New Jersey, USA | Lahore, PAK | Dubai, UAE


(855) 833 7775 (703) 646 – 3043

Please follow and like us:



ClayDesk provides e-Discovery and cloud computing services worldwide.As a Microsoft Cloud Partner, our certified team can deliver innovative and cost effective solutions. We also provide complete document management services.

Visit us at


Please follow and like us:

What if a Naval Officer commits a ‘Civil’ offence?

This is a common question being asked by personnel most of the time. First, a Civil offense in the eys of Naval law is basically a criminal offense. Keeping that in mind, if a serving officer commits a civil offense outside the jurisdiction of Naval unit, he is basically held responsible for his actions like any other citizen would. Navy, however, has the say in determination of jurisdiction for the trial. In other words, Navy could call the entire matter and investigate the matter itself before reaching any conclusions. The magistrate in civil courts is bound to provide all documentary evidence as well as allow the individual to be taken into Naval custody. Subsequently, Navy can then conduct appropriate trial based on evidence and investigation. The JAG attorneys are responsible for handling the case from start till the culmination of the said matter.

Please follow and like us: