More from Gartner Data & Analytics Summit 2024

Solution Architect Nischay Thapa shares his technical highlights from the 2024 Data and Analytics Summit.

How to Make Your Data AI-Ready and Why It Matters

Preparing data for AI is crucial for leveraging the full potential of AI technologies. The emphasis on data governance, continuous quality assessment, and the alignment of data management practices with AI needs highlights the evolving landscape of data strategies. Organizations must prioritize these changes to stay competitive and maximize the benefits of AI initiatives.

AI-Ready Data is Critical: Ensuring data is AI-ready is fundamental to the success of AI initiatives. This involves proper alignment, continuous qualification, and contextual governance of data.
Evolving Practices: Traditional data management practices need to be updated to meet the demands of AI. This includes implementing new data concepts and improving data sharing and governance.
Business Outcomes: AI-readiness directly correlates with improved business outcomes, making it a strategic priority for organizations.
Actionable Steps: Organizations should focus on developing metadata management, enhancing data quality, and forming cross-functional teams to support AI initiatives.
Addressing Misconceptions: Clearing up common misconceptions about AI-readiness can help organizations better prepare their data and infrastructure for AI deployment.

Five things to do to get your data AI ready.

Make data management an AI strategic priority.
The traditional “single version of the truth” is biased – especially true with AI — which learns bias and discernment.
Develop metadata management, data observability and D&A governance to support your AI program.
Put in place cross-functional teams to support your AI-ready data requirements.
Learn and expand AI uses of data by leveraging metadata that comes out of AI processes continuously.

What to put in place now, for AI success in the future

Organizations must evolve their data and analytics strategies to keep pace with technological advancements and the growing complexity of data environments. Embracing AI, managing complexity, ensuring trust, and empowering employees are critical to achieving business success in 2024 and beyond. The emphasis on practical actions and governance underscores the need for a structured yet flexible approach to data management.

AI and Business Integration: Generative AI and other advanced technologies will have a profound impact on industries, necessitating a shift from “good enough” data practices to those that can support critical business decisions.
Managing Complexity: Organizations must adopt new methods to manage the complexity of their data environments, using design thinking, process mining, and AI-enabled systems to streamline operations and improve decision-making.
Building Trust: Establishing trust in data through robust governance, observability, and explainability practices is essential to counter the growing distrust in single sources of truth.
Empowering Employees: Enhancing employee capabilities through AI literacy programs, decision-making tools, and dedicated innovation time can lead to better business outcomes and reduced burnout.
Proactive Actions: Immediate steps such as adopting FinOps, establishing D&A franchises, and aligning stakeholder goals with ethical data governance are crucial for staying ahead of the trends.

Here’s our key takeaways for evolving your data and analytics environment.

Prove the value to the organization.
- Use the enterprise value equation to create a clear link between D&A capabilities and business value.
- Optimize resource allocation decisions with a consistent approach to scoring investment opportunities.
- Include relevant intangibles, such as sustainability or DEI goals.
Use FinOps to flatten the cost curve
- Establish a formal, centralized FinOps practice to set and enforce standards.
- Track and attribute costs by using resource tagging and targeted FinOps tools and practices.
- Conduct regular vendor price/performance reviews to assess the health of your environment.
- Seek out vendors offering augmented FinOps capabilities to progress your maturity and automation aspirations.
Establish D&A franchises.
- Identify best practices in skills, processes, and technology across the organization.
- Use those best practices as a template for others to leverage.
- Consider establishing some franchisees as designated thought leaders.
Balance your complexity
- Avoid the temptation to try to fix problems by adding complexity.
- Simplify the governance process with policies, not rulebooks.
- Use design thinking principles to simplify your technology environment.
- Consider using D&A-enabled complexity as a source of competitive advantage.
Understand your ecosystem.
- Invest in systems that capture contextual data.
- Create a digital twin of your products, processes, organization, or environment.
- Invest in process mining tools to discover, monitor, and improve business operations and processes.
- Use AI to build automated responses to complex phenomena.
Augment with AI-enabled systems
- Invest in augmented data management tools, for established use cases such as data catalogues and migration and to support advanced use cases like active metadata.
- Boost the efficiency of business analysts with augmented pattern discovery and communication capabilities.
- Embrace decision automation for more use cases with self-learning AI systems.
- Invest in augmented analytics capabilities such as NLP/NLG to reach emerging groups of users.

Practical Composable Analytics? What you need to know

There is a growing need for composable analytics to adapt to the changing demands of businesses. By shifting from rigid, IT-driven models to flexible, business-composed analytics systems, organizations can achieve greater agility and responsiveness. The integration of natural language processing and establishing fusion teams highlight the importance of collaboration and advanced technologies in driving analytics innovation.

Composable Analytics are non-negotiable: Composable analytics are essential to remain competitive and agile. Leveraging modular, low-code, and AI-driven technologies to create adaptable analytics solutions.
Four Essential Steps: Developing business technologists, building modular capabilities, assessing composability, and curating the analytics experience are crucial steps for implementing composable analytics.
Flexibility and Innovation: Organizations must equally value technical and commercial flexibility, allowing users to consume and purchase products in ways that best suit their needs.
Semantic Layers and Natural Language rule: Implementing a stand-alone semantic layer centralizes business logic and metrics, while natural language processing enhances the context and usability of analytics content.
Collaboration is key: Fusion teams comprising business and technology professionals will drive innovation and create the best analytics capabilities.

Five steps to bringing D&A to the centre of your business.

Prepare – Build a joint team of application developers, citizen developers and business analysts to add analytics capabilities to applications through agile assembly and reassembly of analytics services.
Act – Pilot composable analytics in the cloud, establishing an analytics marketplace/exchange to enable collaboration.
Deliver – Leverage composable analytics to drive innovation by incorporating advanced DS/ML capabilities into analytics applications.
Reuse – Establish a marketplace/exchange to save data and analytics packaged business capabilities (PBCs).
Grow – Extend analytics capabilities by establishing composability within existing D&A portfolios.

What is the REAL cost? Measuring and Quantifying Cost, Risk & Value of AI

Past mistakes and the complexity and variability of costs associated with AI and GenAI initiatives, highlights the need for detailed cost modelling and risk assessments. By categorizing AI opportunities and employing comprehensive cost and risk management strategies, organizations can better navigate the financial and operational challenges of AI implementation.

Cost Management – Effective cost estimation and management are crucial for the success of AI initiatives. Organizations must account for various cost components and prepare for significant cost variability.

Risk Assessment – Identifying and mitigating risks related to AI, such as data quality and model security, are essential to ensure reliable and ethical AI solutions.

Strategic Categorization – Categorizing AI initiatives into Defend, Extend, and Upend helps in targeting investments and aligning them with business goals and risk profiles.

Utilizing AI Monitoring Tools – Leveraging AI risk monitoring tools can provide insights into model performance, fairness, and security, thereby enhancing the reliability of AI systems.

Quantifying Benefits – Organizations should measure multiple aspects of AI benefits, including productivity, customer experience, and product improvement, to fully realize the value of AI investments.

Five top tips on how to manage costs in this brave new world.

Effectively estimate cost, risk and value by categorizing your opportunities into three initiative types (DEFEND, EXTEND, UPEND).
Build a cost estimate by focusing on the cost factors relevant to your initiative type.
Mitigate risk by allocating budget, resources, training and tools to ensure a safe, secure, ethical and reliable GenAI solution.
Predict your GenAI cost and cost risk by modelling several scenarios incorporating over as many cost line items as possible.
Maximize value realization by incorporating multiple benefit measures of GenAI (not just productivity).

Emerging Practices for Decision Intelligence: The Next Leap for Data, Analytics and AI

Decision Intelligence(DI) is extremely important in transforming how your organisation makes decisions. By integrating advanced data analytics, AI, and process modelling, DI provides a structured approach to improving decision accuracy, speed, and impact.

Decision Intelligence is a differentiator: Adopting DI can significantly enhance an organization’s decision-making capabilities, providing a competitive advantage in a rapidly changing business environment.
It must be structured: Implementing DI requires a systematic approach, from prioritizing decisions to embedding and monitoring them within business processes.
Integrate Data, Opinions, Analytics, and AI: Effective DI integrates various elements, including real-time data, human insights, analytics, and AI, to support, augment, and automate decisions.
Stop, collaborate and listen: Building trust and collaboration between different stakeholders and decision engineers are crucial for the success of DI initiatives.
Continuously Improve: Continuous monitoring and evaluation of decisions ensures that the decision-making processes remain adaptive and aligned with organizational goals.

Data Mesh vs Data Fabric? Identify the Benefits and Risks Before Investing

It is important to understand both data mesh and data fabric before making investment decisions. This means you need strong metadata management and governance practices for successful implementation:

Informed Decision Making: Organizations need to carefully evaluate their metadata and governance maturity before choosing between data mesh and data fabric.

Metadata Management: Strong metadata management is crucial for the success of both data fabric and data mesh implementations.

Balancing Priorities: Data fabric focuses on metadata and technology, while data mesh emphasizes organizational models and practices. Balancing these priorities is essential for effective data management.

Avoiding Silos: Avoid returning to fragmented data management by improving metadata and governance maturity before adopting a decentralized approach like data mesh.

Continuous Learning: Learn from early adopters and continuously refine data management practices to achieve better alignment with business goals and data management efficiency.

Three ways to ensure a successful investment in AI.

Start with existing Logical Data Warehouse architecture, then branch into fabric, evaluate mesh, or combine both approaches.
Ensure metadata management practices are in order before decentralizing data management.
Avoid the proliferation of data mart silos by improving metadata and governance maturity.

Upcoming Webinar

How to optimise AI in your business

A discussion with Ignite’s Consulting Services Manager Kris Hunter and Solution Architect Nischay Thapa on how to translate Gartner’s trends into practical, business value with Ignite.