Welcome to New Customers

We are pleased to welcome Adams & Reese, Gibbs & Bruns, Morgan Lewis & Bockius, Stafford Rosenbaum, Cokinos Bosien & Young, Westlake Chemical, and AIG as new customers using Disco software either through us or through one of our channel partners.

We are currently hiring sales and marketing staff to spread the word about Disco to law firms, corporate litigants, and potential channel partners. We are also meeting with channel partners in major U.S. cities with the goal of forming relationships with 2 – 3 channel partners per major city by the end of 2013. We are looking for a mix of channel partners, from established players in the e-discovery space to partners with strengths in other legal-service areas (collection, copying, forensics, etc.) who have strong relationships with litigation partners and want to enter this space.

If you are interested in working with us, email CeCe Cohen.

Lawyer Matching Services

A popular kind of legal startup in Silicon Valley is the lawyer matching service. See, for example, Law Pivot, Rocket LawyerLawdingo, and Avvo.

The basic idea is that lawyers are difficult to search for and expensive to try, but many lawyers are actively searching for business and are willing to spend money and time to meet new clients. These services sign up lawyers and charge them or get them to provide free or discounted legal content, either to the site or to potential clients one on one; and, in exchange, the services send potential clients to the lawyers. For clients, these sites provide a centralized place to find lawyers, discounted prices, an easy way to schedule appointments and compare advice and qualifications, and, eventually, reviews from clients or others.

The tech press likes these startups; traditional lawyers don’t. Why?

Finding a good lawyer is a real problem. Too many small to medium sized businesses ($100M a year or less) subsist on bad legal advice because they can’t tell the difference, aren’t willing to pay the price for good advice, or don’t know that they have to pay the price to get good advice; the problem is even worse for individuals with small civil or criminal matters; whether they get a good lawyer is basically a matter of chance.

A single place where these clients could go to find good lawyers would be a great service; it would add value both to the lawyers who used it and to the clients it served. But the main service of that single place is not merely collecting lawyers, reducing their rates, or sharing information among clients about them — the main service of that single place is screening good lawyers from bad, making lawyers better than they are, giving lawyers the resources they need to do great work, and making it easy to assemble multidisciplinary teams of lawyers for problems that require different areas of expertise.

That organization is the law firm, at least as great law firms used to be. Until these startup marketplaces do a good job at screening for quality — something that can only be done by lawyers, or by technology that understands enough of legal practice to distinguish good lawyering from bad — they will be doing a disservice to both lawyers (whose participation is, indeed, a sign of desperation) and to clients (who are not getting what they really need, which is, at minimum, a quality screen). I very much hope these startups eventually develop something like this, what might be described as a parceling out of the marketing and quality-control services that law firms once provided to clients and lawyers; but until they do, they have misunderstood the problem they are trying to solve.

A few words about quality control for nonlawyers: checking for bar membership and bar discipline is nowhere near enough. Unfortunately, in the United States, it is easy to become a lawyer and there is little or no postgraduate specialization or training as there is for doctors. The kinds of things that can be routinely checked are not adequate indicia of quality. And clients can judge bedside manner, but not the quality of the surgery; unlike real surgery, the results of bad lawyering may not show for years in litigation and longer in transactions. This is why the quality-control function of law firms is so important, because, until better technology is developed, only senior lawyers who are themselves good can tell what lawyers are good and can teach them how to be better.

Technology to determine the quality of lawyers — now that would be something!

The Vision

We believe in great lawyers at Disco.

Great lawyers are great because of the original analysis, responsible consultation, and personal advocacy that only they can do. We are the company that will build the tools that will let great lawyers concentrate on these things only. We will automate the rest.

Our first product, Disco, is a 10x faster, 10x cheaper e-discovery document-review platform that lets lawyers find evidence faster. Our goal for Disco is to be, for finding evidence, what Westlaw is for finding law. In what we call “version 1,” the version that is currently available, we’ve focused on making the things that lawyers do 90% of the time as fast and easy to use as possible and on making Disco available on a fair, easy-to-budget-for flat-fee basis that brings complete predictability to e-discovery processing, software, and production costs. This lets lawyers do the work of document review as quickly and happily as possible. In “version 2,” which we are working on now and which we anticipate releasing in fall 2013, we will be introducing tools that multiply the productivity of lawyers by automating as much of the actual review as possible.

The law is a great and noble calling, but so much of what lawyers spend their time on is only a distraction from practicing it well. The great innovations to come in law will be innovations that reduce the amount of time lawyers spend on distractions and increase the amount of time lawyers spend on real lawyering; this will magnify the output of great lawyers the same way power tools magnify the output of a carpenter. The view people often have of lawyers is of paper pushers fighting for anything (or nothing) who don’t understand the ends to which they put the law to use or who are simply inept at using the law at all. But law done greatly is not that at all, it is the study and practice of organizing men and institutions to achieve great things — great deals, great organizations, great businesses, great governments.

I urge you to reach out to us if you want to be a part of an organization dedicated to helping lawyers practice law in this sense. The problems you work on won’t ever be unimportant.

Speed Tests

The number one problem we had with e-discovery software before Disco was how intolerably slow the software was. It would take tens of seconds to run searches, seconds to navigate between pages of results, and seconds to navigate between documents (or even pages of documents).

The obvious cost of this is the time that lawyers spend doing nothing but waiting for searches to run and documents to be displayed. The less obvious cost — but the more important one by far — is the frustration this causes. This is a large part of what makes document review so unpleasant, or what made it so unpleasant in the past.

In building Disco, one of our primary goals was making the things that lawyers do 90% of the time in e-discovery software as fast as possible. That means making searches, viewing documents, and navigating through search results fast, ideally instantaneous. We are proud of what we’ve achieved in this regard.

For example, LexisNexis Concordance recently circulated an ad to the Above the Law mailing list touting the speed of a yet-to-be-released new version of its software. The advertised speeds were 4 seconds for complex searches and 1-2 seconds for navigation between documents. Disco is literally 10x faster and has been 10x faster since we launched the product: complex searches take 1/3 of a second, and document navigation takes 1/10 of a second.


Lots of technical know how goes into getting these kinds of speed. For example, Disco leverages Lucene search technology and a sharded document database (RavenDB) — open source projects to which our engineers have contributed and that they have extended. Lucene is the same technology that drives search at companies like Amazon, Apple, and IBM. To take another example, Disco renders all documents to PDF — the image format that is displayed in the review window — on ingest, so that no time is spent on conversion during the review itself. The PDF documents are then rendered using pdf.js, a state of the art library for rendering PDFs in the browser itself rather than in a plugin document viewer.

Ingest speed and production speed are another area in which Disco excels. Disco’s ingestion and production are fully parallelized, which means we can make them arbitrarily fast by adding an arbitrary number of computers to work on the data. If a client gives us 2.5 TB (2,500 GB) to ingest, for example, we can make that ingestion arbitrarily fast by assigning more computers to work on the job. Each machine takes a part of the 2.5 TB and ingests it. The tasks are centrally coordinated and the information generated from ingest — e.g., email conversations and the like — are recombined to form the final search index. This is another example of technology being used to achieve results that make lawyers happy. By adding machines, we can meet any deadline.

Building great e-discovery software requires knowing what lawyers want, what the pain points are — in this case, speed — and having engineers with the talent and technical chops to address exactly those pain points. Building great e-discovery software also requires focus: the focus not to be distracted from the pain points into building things that are technically cool but that don’t contribute to lawyers’ happiness.

Disco is great engineering put to the service of making lawyers happy. And it is 10x faster and 10x cheaper than our leading competitors, according to their own numbers.