Kiersted

Saturday, May 19th, 2012 Radically Efficient e-Discovery

PROBLEM

A longstanding Kiersted client, one of the largest law firms in the United States, once again came to us to help them emerge from a potentially disastrous situation for one of their cases. They had delivered a huge volume of electronic data to a local electronic discovery vendor that the client trusted. The trusted local vendor agreed to complete the project within seven days, which was an absolute requirement. After three days of processing, the law firm received unwelcome news that the seven-day deadline would not be met, and the trusted local vendor estimated that the files would produce approximately 12 million pages. The trusted local vendor now deemed the client’s initial request as “impossible” and predicted that the actual time needed to complete the project would be between 14 and 21 days.

Desperately in need of expert help and a partner with tremendous capacity to meet its unwavering deadlines, the law firm approached Kiersted. We were aware that the local vendor had failed in its mission, but nevertheless guaranteed that the project would be accomplished on time.

The fact that our client had 215 gigabytes of native files produced from another vendor complicated the situation because the documents had already been processed and were no longer in their original native format. We received native files that were named using their Bates numbers, and corresponding metadata that was delivered in a Concordance database. Our task was to convert the files to tiff format, assign new unique Bates numbers, brand the images, output the images and create an updated Concordance load file. And we had only seven days to do it!

Another complicating factor was the requirement to number sequentially across all previously numbered documents.  As a result, no documents could be numbered until all had been processed. Our client also requested two copies of the ultimate output.

In addition, as part of their privilege review, the law firm would identify documents that needed to be removed from the population.

PROCESS

On the first day of the project (just prior to Labor Day weekend), the hard drive arrived and the Kiersted team sprang into action. We uploaded 305,958 files to our network, swept and validated them. We began converting the files to tiff format only eight hours after we received approval to begin. By the end of the first day we had already generated two million images. During that time, we extracted all the metadata from the Concordance database for use in our system.

Conversion continued on the second day along with quality control on the images. By that evening, more than eight million images had been created. By the end of the third day, the first round of “tiffing” (when all files are sent through the tiff engine) was finished, and the total number of generated pages exceeded 12 million. Complications arose on the fourth day as we prepared to create the final deliverable (number, brand images, etc.). During preparation for numbering, we determined that the previous vendor that processed the native files had failed to extract all attachments from email messages. We alerted the client and after extensive analysis on all email messages, we found 10,410 missing attachments that the client ultimately wanted extracted and added to the production.

The fifth day was a national holiday, yet Kiersted finished “tiffing” and QCing the newly extracted attachments and performed a final round of quality control to ensure that all documents had been fully processed. A total of 8,777,765 images needed to be branded and produced. Our thorough QC process eliminated millions of unwanted blank pages and we began the process to number and brand the images. The branding effort continued on the sixth day while text files were output, and the load files were created. The load file had to be custom built to incorporate the original metadata delivered in the original Concordance load file. Kiersted obtained metadata from the Concordance database and added new metadata for newly extracted attachments.

By the final day, a total of 115 volumes of data had been output and validated; all image and Concordance load files had been created and verified; and we began to copy the data onto hard drives for delivery. The project was completed and Kiersted met the client’s deadline.

RESULTS

Processing such a huge mountain of data from another e-Discovery vendor’s system involves difficulties that are unknown at the onset. Challenges included the need to re-use metadata from a set of post-processed data; identification and extraction of missing attachments; and custom output requirements. In addition, the numbering schema (which had to be sequential and accomplished in a specific order) could not even begin until all files were processed.

The previous vendor’s failure could have created a disaster for our client but Kiersted’s team stepped up, worked 24/7 with great determination over the holiday weekend, and delivered exactly what the client wanted – a high quality product, on time. Simultaneously, we also met all of Kiersted’s other clients’ deadlines.

The commitment of our team to meet our clients’ needs is the foundation of Kiersted’s reputation. Our clients know that when they find themselves in a pinch, they can call us and we will make it happen. In this particular case, our client is extremely gratified that Kiersted accomplished what others had deemed impossible.