Scaling software delivery for Relativity, the market leader of eDiscovery

Industry and company overview

Discovery is the formal legal process of exchanging relevant information and evidence between the parties involved in litigation. Previously parties exchanged paper documents, often exceeding hundreds of boxes and tens of thousands of documents. As the legal industry moved into the digital age, paper documents were replaced by electronic content - email, social media, cell phone communications etc., collectively described as electronically stored information (ESI). The process of collecting ESI is now known as electronic discovery.

Relativity is a complete eDiscovery platform that helps legal teams solve complex data problems during litigation, investigation, and compliance projects. It comes with a set of tools executing on each step of the process on-premise and in the cloud. It also integrates with 3rd party solutions, allowing the creation of custom workflows. It allows multiple users to gather, access, and analyze this data from different sources, thereby streamlining the legal process. Post review, the software automates the production process in which non-privileged, relevant and responsive data is exchanged by the parties involved. This produced data can be easily reviewed by legal professionals to help them find the true story behind complex data and determine case strategies - often saving millions of dollars in legal fees as well as accelerating discovery sixfold, for example, five days instead of a month for review.

More than 13,000 organizations globally, including 198 of the top American law firms, corporations, and governments, benefit from Relativity's advanced analytics and machine learning capabilities.

For the last couple of years, Devbridge has been a partner working together with Relativity to design and build new versions of the product.

Working with complex, unstructured data

Devbridge has been working with Relativity on the product roadmap of the Review stage in the eDiscovery workflow - a stage where all the data is structured and ready for analysis. Since a single case can contain millions of documents, the application needs to be able to load a document list very quickly - getting a response from the API in the shortest amount of time possible and efficiently rendering data on the front-end. Additionally, the same needs to happen for visualization of data in charts as well as processing complex filter queries constructed by users. Large data sets rendered inside the browser introduce complexities into the design of the application - from initial load performance, to on-the-fly manipulation of records on screen.

The unique technical challenge is creating code that has to be capable of handling choice fields with thousands of objects, hierarchical structures of virtually unlimited levels of depth, and record indices exceeding two billion records. Additionally, the application needs to be able to handle a request to export all documents in a workspace, aggregate and visualize data stored in a distributed database.

With the amount of data that needs to be examined with each new case increasing exponentially, the application requires a unique set of design and engineering skills to produce a tool that is responsive, intuitive to use for the user, as well as capable to perform under extreme datasets. It would be physically impossible for users to manually scan all documents included in a case, hence the analytics tools provided within the product allow automation of the review process through machine learning utilization as well as advanced search and indexing algorithms.

Furthermore, Relativity is an extensible platform, used by 3rd party developers (such as Devbridge Group) to create unique client solutions. Delivery of the core product roadmap requires the understanding and technical discipline to avoid any breaking changes that might affect existing customizations and extensions.

Agile at scale

As a global leader in the e-discovery space, Relativity excels at addressing users’ needs by building nimble, autonomous teams who take ownership of key areas of the product. A like-minded philosophy and working model allow Devbridge to extend Relativity’s product delivery bandwidth with multiple teams, contributing working software to the companies release train. Creating an on-demand, variable delivery capacity is a value-add Devbridge provides for software companies in a growth stage.

Why is this important? Well, scaling internal teams undoubtedly puts pressure on teams who are often asked to contribute to interviewing, onboarding and knowledge transfer. In many cases, high functioning teams are split to ensure new team member have the regular support of veterans. This can be a lengthy process that impacts the team over several sprints (or several months) ultimately impacting the team's ability to focus on product delivery. When a deadline looms, finding a way to minimize this opportunity cost can help the team to better support the company’s strategic goals.

The relationship between Devbridge and Relativity is built on seamless integration - operating Devbridge and Relativity teams as a single organization, integrating into Scrum of Scrums, and adhering to the same internal release cadence. Our objective is to provide the client with domain-specific teams that are highly scalable and variable, depending on seasonality and specific release demands.

Early investments in building domain knowledge and ongoing transparency in communications and delivery are key to our successful partnership. For example:

  • Devbridge provides cross-functional product delivery teams including engineering, test, design and product management professionals who are experienced at launching product to market.

  • Much like Relativity’s employees, new Devbridge team members participate in product training and business orientation activities as part of their onboarding to Relativity to provide context which guides discovery and development.

  • Teams use shared tools for backlog management to increase visibility to inter-team dependencies.

  • Devbridge teams are accountable to the same code standards and working agreements including quality, performance, test coverage, documentation and accessibility requirements.

  • Devbridge teams follow the same delivery cadence as all other teams and contribute to all cross-team ceremonies including bi-weekly product demos and milestone reviews

  • Devbridge and Relativity conduct quarterly partnership retrospectives to ensure challenges are quickly identified and we continuously improve our working relationship.

As Relativity uplevels their own tools and practices Devbridge teams are active collaborators. Originally tasked with modernizing the UI framework for Relativity - moving away from ASP.NET Webforms using modern Javascript libraries -- Devbridge took ownership of the surrounding layers including updating and optimizing related APIs, removing business logic from the presentation layer, creating a framework agnostic component library and building new test automation infrastructure. By replacing legacy robot framework with Selenium, Devbridge created automated testing tools which were more accessible to a majority Relativity’s in-house engineers driving broader and more rapid implementation of automated tests among other teams.

Scaling software delivery for Relativity, the market leader of eDiscovery

Summary

Growth driven by successful adoption and retention is one of many factors that present product companies with pressure to scale delivery teams. Competing priorities driven by business events, regulatory changes or the tension between customer commitments and new business opportunities are familiar challenges to every product team (software companies with an established SDLC and high performing teams are no exception.) Devbridge provides domain-specific cross-functional product teams that can scale up in a manner of weeks and help mature software companies diversify knowledge retention as well as address peak throughput demands.

Increasing operational efficiency and processing with e-discovery workflows

Increasing operational efficiency and processing with e-discovery workflows

Read case study