Unleashing the data potential is only possible with the right data integration tools. David Lipowitz, Solution Architect Team Lead at Matillion details how remote, distributed data teams can use enterprise data in real-time with the best tool.
With the worldwide boom in remote workingOpens a new window , companies are quickly taking steps to keep critical work on track and prepare for potential business disruption. They need to practice social distancing, actively redefine distributed data strategyOpens a new window , and challenge data architects to balance a large range of circumstances. Now more than ever, businesses need to analyze trends and data, gain insights, and make better decisions. Today’s data teams need tools and technologies that enable them to increase efficiency,Opens a new window work across distributed teams, collaborate in real-time, and be ready for any business disruption â€“ including a stay-at-home order from the government.
Data teams should seek data integration products that easily support distributed, collaborative work and disparate teams. In a remote environment, development teams Opens a new window and individuals are working from a variety of locations and time zones. To support them, an enterprise will want to maintain the integrity and consistency of data projects to produce business insights faster. Development and data analysis teams must be able to work together, in real-time, to easily update existing models and create new ones. A remote work environment requires solutions that enable and foster collaborationOpens a new window .
A data integration tool should include the following attributes to maximize efficiencies:
1. An Intuitive, Browser-Based User Interface
A single shared user interface, procured and provisioned in the cloud, is a collaboration Opens a new window cornerstone that is accessible from anywhere, avoids local configuration and maintenance, and continually adds features and functionality that support highly collaborative teamsOpens a new window .
Tools that are locally managed are difficult to maintain and configure, and they present a variety of logistical problems. Take for example, the â€œit works on my machineâ€-effect, which is difficult to troubleshoot when different users have different versions of locally installed software and experience different behavior. Even worse, these issues can spread throughout the team, entangling additional developers and taking them away from meaningful work. Centrally managed software ensures the whole team is on the same page.
One of the opportunities a browser-based interface affords is the ability for team Opens a new window members to co-develop the same workflowOpens a new window , in real-time. Think Google Docs for a data team. This enables team members to see the effects of adding or modifying components instantaneously (and vice-versa). Multiple team members can work on a single workflow, allowing each to focus on their respective areas of expertise. In doing so, a project can be completed, tested, and delivered more quickly than a single resource could manage. This real-time coordination also allows team members to perform code-review tasks on the fly as a workflow is being built, saving time that would otherwise be spent providing context to a reviewer. The end result is increased speed to delivery with the potential to increase quality through paired programming techniques that reduce bugs through real-time code reviews.
2. Fine-grained Permissions
Data teams need to work with solutions that offer flexible user and permission management, and that has the ability to create unique and specific roles for business analysts, project managers, and product owners. It is important for developers, engineers and their counterparts to look at the same workflows and speak the same language, without the worry that less technical resources can see, modify, or execute something that could have negative impacts.
Auto-documentation is an enormous time-saver, even without going into its collaborative benefitsOpens a new window . Maintaining documentation is extremely important for most organizations, but it is also an incredible drain on overall throughput. At times, the lifespan of a document version can be calculated in minutes, so keeping up with iterations requires dedicated resources. Unfortunately, in many cases, tracking changes falls by the wayside. Developers should seek products that make it easy to follow updates and offer thorough, hyperlinked, and clearly formatted documentation Opens a new window to align developers and stakeholders.
Even with a solution’s advanced permission control features, sometimes you don’t want to give read-only access to stakeholders. In those cases, you can give them documentation instead, which helps answer their questions and promotes alignment on the business rules developers implement.
4. Git Integration
Git is the go-to version control system for today’s developer, with its feature-rich support of distributed repositories, easy branching and merging, and straightforward workflows that enable distributed teams to contribute to the same codebase in a managed way. Companies should look for tools that allow their developersOpens a new window to track the history of their own changes, and merge those changes into centralized branches along with other collaborators. Ideally, your chosen solution will allow your entire team to contribute changes but we don’t want those changes ending up in production before they’re tested. Git allows you to control when your changes are moved from development to QA environments and when they go from QA to production. Seek a solution that allows easy management of your job development and promotion processes.
Undoubtedly, collaboration is hard to master for remote teams. Fortunately, there are solutions that can reduce the challenges teams face, enabling them to stay connected, efficient, and productive. The key is to take advantage of tools Opens a new window that are well suited to the demands of co-located team members from a wide variety of skill sets and functions and take into account the types of problems faced by modern data teams.