James Maguire, eWeek’s Editor-in-Chief, moderated a discussion in December on the future of analytics, which included some intriguing 2023 predictions. A good panel includes different and unique perspectives and this one didn’t disappoint. The panelists were:
- Radhika Krishnan, Chief Product Officer, Hitachi Vantara
- Torsten Grabs, Director of Product Management, Snowflake
- Krishna Subramanian, Chief Operating Officer, Komprise
- Barry McCardel, Chief Executive Officer, Hex Technologies
In 2023, the panelists focused on a few key data analytics themes:
Actionable data insights:
There will be a greater ability to connect data and insights to actions and decisions. Data collaboration and sharing will extend to the logic that knows how to derive insights from the data. Tighter feedback loops will be created between decision makers and data teams.
Cybersecurity continues to be a top concern for IT organizations and CXOs as tactics become more complex, intelligent and destructive from foreign actors. As a result, panelists expect a heightened focus on data governance, data security and data privacy from enterprise data teams. Smart data workflows will efficiently enable data workflows across the edge, data center and cloud infrastructure and across organizations/departments to meet compliance and security needs for different data sets.
Machine learning maturity:
The panelists predict that we’ll see a rise in marketplaces for sharing ML models for re-use, which will help demonstrate the impact of data science projects/investments and require less expertise in the core ML technology. Reusable models should reduce the cost of compute and barriers to entry.
The major cloud providers offer these already; expect more startups to compete this year.
Edge data management:
There will be a greater need to collect and analyze data at the edge, where there is exponential data growth from sensors/IoT and mobile apps.
Beyond 2023: Here’s what panelists predict for data and analytics strategies in the next 10 years:
Machine learning everywhere:
ML will become part of the decision-making process and daily workflows, as the models become easier to build and the tools become easier to use for the average information worker. Like what happened with self-service in the business intelligence (BI) market, imagine if you don’t have to be a data scientist to work with advanced analytics and ML applications? However, other challenges remain, including how to clean up, organize and contextualize data effectively across edge and hybrid infrastructure to enable smart cities, smart buildings, electric vehicles and more.
Sustainability gets real:
In the era of “data hoarding,” when you think about the edge alone, there’s too much data being collected and not enough space. Intelligent extraction and curation strategies to manage and keep only the data that are needed will emerge to streamline data management, cut costs and conserve energy. How to set, measure and track sustainability goals will become a priority.
Ethics of data analytics:
If we rely too much on ML, we might be missing ethical and social context. This remains a big unknown. Explainability in analytics and AI is gaining traction.
Natural language AI technologies like GTP-3 –ChatGPT is the over-hyped face of this–will change creative workflows across the board with wide-ranging impacts, including upskilling and empowering people to do more creative work and democratizing data science. This trend is happening faster than we think and will need to be on the radar of anyone who works in a data management or data science/analysis role.
I tried to capture the key points from the panel. It was only 30 minutes, but full of great insights. Credit to James McGuire for driving great engagement and keeping the conversation flowing. You can check out other podcasts and posts from James here: https://www.eweek.com/author/jmaguire