The COVID-19 pandemic has forced companies to shift to remote working. While most companies switched to work from home within a matter of weeks, companies working on complex technical skills like Data Science had a tough time. A report by AIM Research suggests that 34% of Data Science professionals had a negative impact on their productivity due to work from home.
Access control
For most data teams, the abrupt shift to remote working led to the biggest challenge of how to access the data. Majority of the data exists on on-premise servers. It is a critical challenge which led to forced cloud migration for most data-driven organisations. Since companies have built distributed remote teams, protecting sensitive data, identifying access management took the most amount of time. The access policies for data scientists had to be changed, made complex and layered.
Collaboration and communication
To ensure data scientist's productivity from a remote location, there need to be supportive Data Science processes and tools. Ad-hoc communication can be challenging at times. Since most data teams are used to working from the same environment, they can even feel lonely and distracted at times. Remote working requires everyone to document every little aspect of working, which is an added burden for data teams. For companies where documentation has remained terrible for years, following the same standards can be frustrating for data science teams.
Documentation
In ML pipeline, many dependencies are on proper documentation. Since companies and their respective teams are not used to function based on what's been documented, it is challenging to collaborate without documentation. Earlier, the teams could discuss the problems in-person. Now they have to get on a formal way of communication to solve small problems. Data scientists many times download their own version of development tools, which requires them to document separately and knowledge transfer to colleagues working on the same project.
Access control
For most data teams, the abrupt shift to remote working led to the biggest challenge of how to access the data. Majority of the data exists on on-premise servers. It is a critical challenge which led to forced cloud migration for most data-driven organisations. Since companies have built distributed remote teams, protecting sensitive data, identifying access management took the most amount of time. The access policies for data scientists had to be changed, made complex and layered.
Collaboration and communication
To ensure data scientist's productivity from a remote location, there need to be supportive Data Science processes and tools. Ad-hoc communication can be challenging at times. Since most data teams are used to working from the same environment, they can even feel lonely and distracted at times. Remote working requires everyone to document every little aspect of working, which is an added burden for data teams. For companies where documentation has remained terrible for years, following the same standards can be frustrating for data science teams.
Documentation
In ML pipeline, many dependencies are on proper documentation. Since companies and their respective teams are not used to function based on what's been documented, it is challenging to collaborate without documentation. Earlier, the teams could discuss the problems in-person. Now they have to get on a formal way of communication to solve small problems. Data scientists many times download their own version of development tools, which requires them to document separately and knowledge transfer to colleagues working on the same project.