When is it safe to re-open an office in a COVID-19 world? This is a charged topic of conversation — but one in which data science can empower your company to make educated decisions that reassure your employees.
Ryan Ries and Mark McQuade both work in Data Science and Engineering at Rackspace. They spend countless hours digging into data, understanding what it tells us, determining trends and figuring out whether the data has predictive or forecasting value. This all has applications in a wider business context, but Ries believes data science has a particularly vital role to play in today’s business climate. “We don’t know what the long-lasting effects of COVID-19 are and so people are worried about going into an office,” he said. “How do you convince them it’s going to be safe? That comes down to data.”
This is the central theme of the latest Cloud Talk podcast. Ries and McQuade join Rackspace Technology CTO Jeff DeVerter for a lively discussion on working with data, cleaning and modeling data, and how forecasting models can help people feel safer.
Using data science to get back to work
In just 30 minutes, the panel explores:
- What data science is and how it helps you make better assessments
- How to make intelligent decisions about the future, based on past data sets
- Adding value and resiliency through cleaning data — and when cleaning should occur
- Best practices regarding building, testing, validating and training data models
- The need to continually refine and retrain models in fast-moving situations
- How open data can help others with their own COVID-19 predictions and forecasting
It might seem like overkill to create your own COVID-19 forecasting models, but there are benefits. “A big issue today is there’s so much misinformation, and so we need to help companies make fact-based decisions,” said McQuade. A recent Rackspace Technology project involved a company affiliated with UC Irvine, which utilized publicly available data, AWS and the power of DeepAR to provide COVID-19 forecasts. It outperformed Institute for Health Metrics and Evaluation (IHME) data. McQuade enthused this was “building reliable forecasting people can actually believe — because the data is real”.
This isn’t a simple thing to accomplish. “You might find it daunting to build your own model,” said Ries, who added that, in order to open safely, businesses must also look at contact tracing to capture people who don’t present any symptoms. But once you get there, this information can become your truth — more than any other source. “The key is to rely on that data as much as possible,” said McQuade. “When executives are deciding when to re-open, they shouldn’t be talking about something they saw on the news. The starting point — your reference for decision making — should always be your data.”