IT-Discovery
Delivered as a service, IT-Discovery helps companies reduce costs in the electronic discovery (e-discovery) process and rapidly solve corporate or regulatory investigations. IT-Discovery provides a level of intelligence for traditional information management systems and archives to create "intelligent repositories". As a result, companies can perform early case assessment, cull-down data more effectively, and find facts almost immediately.
Think of any advanced search page: you have dates to enter, metadata fields to choose, key words to input, filters to apply or not, people or custodians to choose etc. We think there's a better way. Instead of being forced to type in a search query, simply "browse" the corpus and let the relevant information find you.
IT-Discovery is designed to look and feel like the opposite of a search page, while giving much deeper results than could be achieved conventionally:
Additional features of IT-Discovery include:
Ease of use
There is good reason why "training sessions" proliferate in the litigation support area. There is an almost complete disregard for intuitive interfaces in many applications, in favor of features that improve demos or just out of sloppiness. IT-Discovery, on the other hand, is extremely easy to use, often requiring nearly no training.
Scalable Architecture
Many applications are useful but don't scale. The great challenge in building IT-Discovery, since it is grounded in machine learning, was making it scale. This has been achieved with the help of recent developments in cloud computing and parallelization platforms. The entire corpus is addressable at sub one second response times if you wish to add enough machines to the process, or scaled back as needed. It is as muscular a product as it is intelligent and easy to use.
IT-Discovery leverages end-to-end virtualization and a distributed computing paradigm to accelerate every aspect of the processing and application delivery infrastructure. This allows us to complete ordinarily time-consuming tasks at unprecedented speed.
By employing load balancing for both the web and application server level, IT-Discovery can effectively host as many users as a case requires. Typical setups support over fifty simultaneous users; on larger cases we can support twenty times that number.
Distributed Text Extraction and Indexing
Processing has historically been a pain point for many, with vendors requiring weeks to load large data sets. On our infrastructure, waiting weeks is no longer necessary. Every processing job runs simultaneously on scores of servers; larger jobs are simply allotted more machines. The result: processing time is practically independent of data volume. Terabytes can be loaded as rapidly as gigabytes.
Machine Learning
Utilizing the latest advances in social network analysis and semantic search, IT-Discovery can automatically categorize documents in datasets of any size – without any pre-processing or manual tuning.
Deriving topic categories in large document collections requires huge amounts of memory and processing power. Our scientists and engineers went to extraordinary lengths to parallelize this task so that it can be run across multiple servers simultaneously. As a result, we now routinely process and categorize millions of documents in hours – rather than days or weeks.
"Internet-scale" Search
At IT.com, we don't believe in reinventing the wheel. That's why we leverage proven open source technologies to deliver a solution that is enterprise-grade, robust, and scalable. When it comes to data processing, we've extended Apache Hadoop, which is used by Fortune 500 companies to process and analyze petabytes of data. Further, our real-time text search and analysis technology is built on top of Apache Solr, which supports indexes of hundreds of millions of documents. In short, our use of open source technology enables IT-Discovery to analyze and search documents far more quickly and reliably than our competitors' proprietary, custom-built solutions.
Enterprise Reliability
IT-Discovery utilizes an infrastructure that is among the most robust possible. Specific attributes of our architecture include:
- Full server virtualization, in which every server in our grid runs on a virtual machine cluster. If a particular physical machine is about to fail, the server is moved to a functioning machine, automatically
- Redundant switches and internet connectivity reduce the likelihood of an outage due to failed network gear
- Multipath storage connectivity ensures that data is always available, even in the event of malfunctioning cables or controllers
- Massive fiber-channel Storage Area Network supporting 500+ terabytes of high-performance storage provides the best reliability and capacity in the industry
- Datacenter-level redundancy: Every critical element at both server and datacenter level is designed to ensure maximum availability. But at some point, every computing environment, no matter how well-designed, may experience an outage (due to human error, natural disaster, or even terrorist attack). Accordingly, our entire hosting environment is replicated to a second, completely independent, datacenter.
Traditional & Per-Unit Pricing
IT.com offers competitive per gigabyte pricing on its data processing and hosting services. We can also simplify the process of buying document review services by offering a flat price per document reviewed — data processing, hosting and first-pass contract attorney review included. Clients enjoy absolute price predictability, with no hidden costs.