Application Architecture Summit

Modern Application Development for Digital Business Success

REGISTER

September 23,
2021

Virtual Summit

Today's developers rarely build applications alone and are almost always working in teams. Not surprisingly, this practice increased during the COVID-19 pandemic as distributed work became the norm.

Collaboration among developers often takes place across many tools, but one of the challenges of working together is communication.  Developers need seamless communication to streamline their development activities. The more developers are able to collaborate around data, the more useful that data becomes. This is especially true for the time series data that is increasingly the heartbeat of an organization.

The Impact of Distributed Development Teams on Collaboration

There were already a lot of distributed developer teams before COVID. Today, there are many, many more.  Luckily, there are many different tools available for engineers to develop collaboratively, including pair programming tools and tools that help developers work together across time and space. Look no further than GItHub for a great example of a platform that enables developers to work together.

The technical capabilities of many of these tools have come a long way. In the past, they were essentially glorified text editors.  Today developers can write code, debug, execute and deploy all through a web browser. As more and more of these tools come online they're gaining broader traction.

The goal of any data platform is to go from raw data-to-knowledge. Evidence shows that collaboration is key because it helps build ‘common context‘ across stakeholders, which in turn helps build better and more impactful teams  – and gets everyone from raw data-to knowledge faster.

With time series data, effective collaboration allows you to communicate around individual data points, More importantly, it provides teams with key knowledge about that data. As an example, you might receive sensor readings that are hard to understand, but a data scientist in Antarctica knows that every day at a certain time an animal walks by and brushes it, giving new meaning and context to the data.

Best Practices for Enabling Teams to Collaborate Around Data

Some of the best practices I lean on when it comes to collaboration include:

• Share early and share often. This is an idea similar to writing code. Ideally, you want to push out smaller and more frequent updates, as opposed to keeping a long-running branch open and then pushing out a lot of code at once.
• Keep it short and sweet. Similarly, smaller and more frequent communication is always preferred. Without it, something might get lost and individuals start operating on their own timelines. But if communication is more frequent, you’re much less likely to do the same work twice.
• Annotate everything when possible. This supports the notion of asynchronous communication. The more you annotate and write things down, the easier it is for people across time zones or with different schedules to continue moving a project forward. Anything you can put in text and share is very useful and provides greater context to those you’re working with.
• Make data easily accessible. Allowing people to view key data without jumping through hoops is vital. Telling people that the data they need is in a data store that they need to connect to and then bring it into their own tools to start working becomes very tedious and time-consuming. The more accessible your data is, the easier it is to collaborate on that information.
• Be responsive when working across time zones. Even a five-minute delay to reply or respond can be detrimental across time zones because that delay is amplified. This is particularly true with teams distributed around the world — even a slight delay could cause communication to bleed into after-work hours, for example. Being extra responsive, even if you’re letting your team know you're looking at or investigating something, is useful.
• Put the right tools in place. In my opinion, a cloud-based platform that allows you to build and annotate processes and data flows for time series data is vital. It means the data is always at your fingertips so you can share dashboards instead of queries, views of info instead of raw access. You can add visualizations or explanatory notes and share them within the organization so anyone at any point in time can see the context you've applied to the data and can continue to layer on top of it.

Translating Collaboration Techniques to the Real World

There are a number of use cases and examples of how time series data analyses and trends benefit from collaboration. Notable examples include understanding the pulse of devices, equipment, customers and software. Among these, the most common ones involve IoT and analyzing sensor data.

Take oil drilling, for example. There might be individuals on the oil rig itself and others on land collaborating on a particular set of data. The team on land has all the oil platforms reporting in, and they see an anomaly across three or four of them. In this case, it might be the person that’s on the rig itself who provides context about what's happening in the real world. Was someone sick? Did they slip on something and inadvertently alter the sensor data?

In this case, there are two individuals with two very different views using tools to share information as quickly as possible. A simple annotation on a dashboard could indicate that an anomaly was created by someone clumsily but harmlessly tripping on a cord, meaning back at HQ they don't need to panic or report it to authorities to begin a long, costly and drawn-out investigation. When the on-land worker clocks in at 8 a.m., they can immediately see and understand what happened with the right collaboration tools and processes in place.

In a nutshell, that’s how collaboration makes time series data more valuable.

The knowledge prevents teams from duplicating work or spending resources on something that isn’t a major issue to begin with. In the end, it makes work more efficient and more productive, which is the goal of any development team.

Russ Savage is Director of Product Management at InfluxData where he focuses on enabling DevOps for teams using InfluxDB and the TICK Stack. He has a background in computer engineering and has been focused on various aspects of enterprise data for the past 10 years. Russ has previously worked at Cask Data, Elastic, Box, and Amazon.