Industry executives and experts share their predictions for 2025. Read them in this 17th annual VMblog.com series exclusive.
By Shanmuga
Sunthar Muniandy, Director of Architecture & Chief Evangelist,
APAC of Denodo
As an eventful 2024 is coming to an end, the year of
AI-powered applications and agents truly made meaningful intellectual
contributions towards business and life. As we move towards 2025 (and beyond),
we are excited and bullish about how AI computing will empower the world in
unlocking knowledge and automate actions through intuitive interfaces and
powerful agents.
Ensuring Governed AI Deployments
As AI-powered applications and agents will play a more
pronounced role in life in 2025, AI governance will take
center stage as a critical focus for businesses and government. As AI becomes
increasingly integrated into government and business operations, industry is
turning towards building robust governance frameworks for ethical, transparent,
and accountable AI. Government and business will need to invest into developing
policies and practices that will address AI data, risks, observability,
security, etc., while ensuring compliance with regulatory requirements.
Defining Solid Data Foundation for AI
At the core of solid AI governance is robust data
governance. As the organization is only as good as its data, an
innovative, agile AI data foundation is critical to enabling
responsible and compliant AI deployment. This AI friendly data foundation
empowers organizations to optimally manage AI augmented data for departmental
AI projects (e.g. AI-driven fraud management,
AI-powered customer engagement, ESG, etc.), enabling business self-service of
trusted enterprise data (enterprise knowledge base) for AI initiatives,
ensuring regulatory compliance while democratizing data access across the
enterprise circumventing the potential hazards of departments barreling ahead
with AI projects without IT’s compliance check while maintaining productivity
levels in a best-of-both-worlds solution.
Decentralized Data Management
After struggling for almost four decades, companies realized that the
traditional data management practices and approaches (replication and
centralization of data or monolithic data management) are not keeping up with
the needs of AI and modern applications. There has been a steady trend with in organizations
moving away from monolithic data management and towards logical data
architectures like logical data fabric and data products, enabling organizations to handle data by
business-critical domains. This approach allows organizational stakeholders a
more flexible, scalable and thereby efficient way of managing enterprise data,
while embracing the distributed data ecosystem that we commonly find in
organizations. An interesting capability in data management that is becoming
mainstream is the incorporation of AI techniques for data management. This
shift will support the growing trend of decentralized AI governance, allowing
for more localized and context-specific oversight of AI systems. We predict
this trend to continue strongly in 2025 and beyond.
Hybrid Cloud Architectures is the Future
Privacy regulations, IT risk mitigation, and cost
optimization considerations are pushing organizations to adopt a hybrid cloud
approach in their data and system architecture, by complimenting their existing
on-premise legacy systems with a mix of public and
private cloud environments to implement different business applications and
services and store different classes of data. An area which is a gap in this
hybrid cloud approach is – data management for hybrid cloud data ecosystem.
This trend, which will have steady growth in 2025 (and beyond), provides
organizations with greater flexibility and control over their data and AI
workloads, enhancing their ability to meet regulatory requirements and optimize
performance. The hybrid model will also support the integration of diverse data
sources, facilitating more comprehensive and accurate AI insights.
Data Observability
The demand for AI transparency has subsequently led to
increased demand for data transparency. Data observability, which allows
organizations to monitor data health, lineage, and usage, will thus become a
standard feature. It will enable companies to clearly trace their data’s source
to ensure accountability before training their AI technology’s large language
models (LLMs) on their data pools.
Embedded Regulatory Capabilities
Regulatory bodies are pushing businesses to focus on data
transparency, especially when it comes to AI-driven decision-making. Data
governance architectures will include frameworks for tracking data provenance,
to ensure explainability when it comes to new data and AI regulation
compliance.
To navigate the complexities of AI governance and
decentralized data management, industry collaboration and standardization will
be essential. Organizations should work together to develop common frameworks,
best practices, and standards that promote interoperability and consistency
across different systems and platforms. Companies that embrace data
infrastructural agility will be better positioned to leverage AI’s full
potential for a competitive business edge, while also complying with stringent
compliance regulations for responsible AI deployment and innovation.
##
ABOUT THE AUTHOR
Shanmuga Sunthar Muniandy (Shan) is the APAC Director of Architecture
& Chief Evangelist at Denodo. A Technologist with a career spanning over 20
years, mostly in Data Management and Analytics, Shan has garnered extensive
experience in the areas of Digital Transformation, Enterprise Data Management,
Enterprise Architecture, Business Advisory, Advanced Analytics Solutions,
Analytics Lifecycle Management, Cloud Architecture, Innovation Centers,
Enterprise Customer Intelligence Management, Enterprise Fraud Management and
Enterprise Risk Management (to name a few).
