researchHQ’s Key Points:
- In the coming year, enterprise tech will evolve by adopting new technologies to solve old challenges.
- The edge will emerge as the new frontier of innovation, improving automation, security and efficiency.
- Organizations will increasingly adopt machine learning solutions without necessarily making substantial investments in data science teams.
- Growing momentum in augmented reality (AR) and virtual reality (VR) will continue to drive Workplace 2.0 Initiatives.
- As ransomware attacks continue to grow in sophistication and frequency, security and data protection will remain an evolving priority.
We all went into 2020 with a plan. Those plans were rendered irrelevant just a few months into the year. Organizations quickly rolled out contingency plans and put non-essential initiatives on hold. This may lead one to believe that 2020 was a wash for technology innovation. I would argue otherwise. In fact, organizations deployed inspired solutions to tackle considerable challenges.
Here are a few observations from 2020 and five enterprise tech predictions for 2021.
1. The Edge Is the New Frontier for Innovation
Amazing things are happening at the edge. We saw that on full display in 2020. Here are a few examples:
- When the pandemic first hit, a lab testing company rolled out 400 mobile testing stations across the United States in a matter of weeks.
- A retailer relocated their entire primary distribution center, which was in a state under stay-at-home orders, to fulfill an influx of e-commerce orders from a new location.
These organizations used existing edge investments to react and innovate with velocity. And in the year ahead, we will continue to see prioritized investment at the edge.
Network reliability and performance directly impacts employee and customer experience. That alone led to expansive SD-WAN rollouts at the edge and in home offices. Simple SaaS-delivered solutions (inclusive of hardware) will further improve security and user experience wherever employees choose to work. And this will start a trend in which these solutions become the norm.
Additionally, I expect organizations to increasingly adopt secure access service edge (SASE) solutions. Legacy network and security architectures create unnecessary hair pinning and performance degradation. Instead, our future will lie in application and infrastructure services that are defined in software and deployed and managed as software updates. While upending legacy procurement processes along the way, organizations will dramatically improve performance and security.
We are also getting far more intelligent at the edge, with the ability learn, react and optimize in real-time. Furthermore, we are seeing new opportunities for infrastructure consolidation at the edge, reducing the number of specialized appliances required to meet technology needs. This is an exciting development as it opens doors for cost-positive solutions where you improve automation, safety and efficiencies, while simultaneously reducing costs.
2. Decentralization of Machine Learning
Staying at the edge for another moment, let’s talk about federated machine learning (FML). We are starting to see early uptake in this area among businesses. Across all industries, organizations are innovating to make better data-driven decisions, while leveraging highly distributed technology footprints.
With compute capacity practically everywhere, federated learning allows organizations to train ML models using local data sets. Open source projects, such as FATE and Kubeflow, are gaining traction. I expect the emergence of intuitive applications on these platforms to further accelerate adoption.
Early ML solutions disproportionately benefited a small percentage of enterprises. These organizations had mature data science practices already in place. ML adoption continues to pick up pace. And that acceleration is driven by turnkey solutions built for “everyone else.” These are enterprises that want to reap the rewards of ML without having to make large investments in data science teams—often a difficult challenge given the industry shortage of data scientists today.
