10 Tips to Manage Data Science Remotely
This article was originally posted on my personal blog
The impact of COVID19 has been more widespread than most of us imagine and has caused many organizations and individuals to adapt to fully remote work.
While remote work has already been studied and proven to reduce the gender and race gap by allowing women and minorities to more efficiently commute to work and deal with childcare, it hasn’t been accepted widely in most corporate cultures.
With many states issuing Stay-at-Home orders about 1/3 of the US workforce is adjusting to nearly 6 weeks of working remotely.
In many organizations, Data Science is a major part of the product or provides the company with insight when making business decisions. We’re needed now more than ever to help our organizations optimize our work and stay productive.
Here are my Top 10 Tips on How Manage Data Science Projects Remotely:
🤓 Stay present in meetings:
The facetime you get with your coworkers is even more important now. Limit distractions by closing other tabs or using the PomoDone or StayFocused chrome extensions. I turn my phone off when I’m in the middle of deep analysis, close all other tabs but zoom for meetings, and take notes on paper to avoid getting distracted when on my browser-based to-do site.
📢 Communicate often:
Don’t wait for a big meeting to bring up your concerns. Most companies use something like Slack or Microsoft Teams to make quick communication feasible. Having an “open zoom” policy can help new employees. For senior employees, it’s important to keep the mentorship and learning opportunities going even though we’re working from home. Allow them to look over your shoulder or guide them through your process in creating features or fitting a model.
✨ Schedule Lunch & Learn time:
Pair programming hasn’t changed much so you can still meet with a colleague one-on-one and work through a problem. Don’t be afraid of cross-training. Data Science folks need to understand the big elements of data engineering as well as telling a story with their visualizations.
☁️ Save it to the Cloud:
No matter how confident we might feel in our high-speed internet, there is no assurance it won’t cut out because of high demand. Save your data, code, source files, all of it to your secure cloud storage. This is key for your colleagues to be able to access your work whether you’re on PTO or if you’re having internet issues.
💾 Save Training Data Locally:
This may seem like the antithesis of #4, but the former makes it easier for your colleagues to have access to your work, this makes it easier for you to train models even if your internet goes down. Download what you can, given your computer’s storage. If you have just your training, testing, and evaluation datasets locally you can do most modeling without having access to the internet.
⏱ Stay Aligned with Goals:
Especially if you’re at a company that wasn’t super open to remote work previously, you need to be more vigilant in tracking your tasks. I can assure you managers will be watching your Jira, Asana, or other Kanban boards more to better understand how you perform remotely. Don’t feel like a task is too small to track. It’s better to have too much detail than a conversation at the end of the quarter about you not doing enough.
☎️ Prepare for a shift in work:
Many Data Scientists that were customer-facing are working more on technical work. A lot of people are also noticing a shift towards internal operations type of projects where you’re tasked in evaluating your companies strengths and opportunities. Explore operations data if given the opportunity and make a big impact by finding areas where the company might be wasting money.
💡 Explore new projects:
Look for out of the box ways to solve the different problems you might be facing. I’m looking at ways to pull in additional data (GPS, video, etc) to give my models more context to find patterns in. Your company may be in the position to help with the pandemic and you can look into logistics, NLP for low-resource languages, and data cleaning for Data Science projects.
I had the pleasure of attending the TWIML webinar on Responsible Data Science in the Fight Against COVID-19 this week and Rob Munro described it best:
If you turn up at a hospital to volunteer, they’re not going to let you into surgery, they’ll give you a mop and bucket and tell you to start sanitizing the floors.
If you want to help with the pandemic you may find yourself cleaning data and not working on the cutting-edge models if you have no experience in epidemiology or disaster relief.
💎 Take extra precautions with model bias mitigation:
We’re living in unprecedented times. Make sure to check for bias in your models by understanding how you sample segments, especially protected classes. If you’re like me and are at a CRA (consumer reporting agency) it’s likely your work can impact who gets a job, who get approved for a loan, or who can get certain types of housing (or rent/mortgage relief). We need to take even more care not to cause harm by perpetuating bias with our data models.
🎆 Level Up:
If you have gaps in your knowledge, whether it be hard statistics or product analytics you may have more time to improve your skills. There are multiple resources offering free or discounted courses including: Pluralsight, Udacity, and DataQuest.
Bonus:
🙏🏾Be Grateful
Only about 1/3 of workers in the United States have the opportunity to work from home. While it may feel weird not seeing our coworkers in person, weekends don’t’ feel the same and we find ourselves consuming more tv and video games than we’d typically like, we are extremely fortunate to have jobs, much fewer ones that allow us to work from home easily. Most of our roles pay very well compared to some of the people most impacted by COVID19. If you can, donate to your local food bank or a restaurant worker’s fund.
