Recently, I had the opportunity to attend a very interesting discussion/webinar organized by PEX Network which revolved around the topic of ‘Using AI to improve Master Data Processes’. The session was hosted by Francesca Di Meglio and featured panelists - Andrew Hayden, Shreyashi Das and Rishu Gandhi, all of whom provided key and valuable insights.

The interactive session also involved a good mix of audience participation (through the infusion of Q&A), and provided key insights and perspectives on:

• The relevance and criticality of ‘Master Data’ for the organizational success.
• The reason for AI becoming the ‘Central Focus’ for enterprises.
• The relevance of managing ‘Master Data’ and why that could be a challenge.
• The possible reasons many enterprises shy away from implementing AI across their systems and processes.
• The safety involved in deploying AI on top of systems, in the context of privacy concerns.
• Ease of implementation of AI in cases of prevalence of poor data quality.
• How AI is helping businesses demonstrate ROI on their data initiatives.
• Benefits of combining the master data management and key automation capabilities for driving enhanced enterprise processes.

Key takeaways from the insightful discussions between the panelists and from the audience Q&A:

• AI is becoming the key focus for enterprises: Since businesses and bottom lines are ultimately getting affected, C-suite executives across organizations are attaching increasing importance to AI. Employees, especially data engineers, are benefitting from AI for solving mundane and repetitive tasks. Large language models (LLMs), ChatGPT and AI powered chatbots are classic examples of AI acting like a trusted confidante in helping solve strenuous tasks.
• Organizations ironically shying away from AI implementation: Due to a lack of understanding about internal enterprise processes, the quality of internal enterprise data is also affected. In this regard, hiring SMEs which work alongside the data engineers could unlock true AI potential.
• Challenges companies face while planning for AI: In planning for AI, companies face pivotal challenges centered on data quality and availability of data, foundational to successful AI implementation. Accessing relevant, high-quality data remains a fundamental hurdle, as inadequate or biased datasets can compromise AI model accuracy and reliability. Additionally, navigating regulatory compliance (especially concerning data privacy and ethical use) demands meticulous considerations to avoid legal repercussions and for maintaining stakeholder trust. Moreover, the scarcity of AI talent capable of translating business needs into actionable AI solutions poses another critical constraint.
• Understanding the safety and privacy concerns in deploying AI atop enterprise systems: This demands meticulous attention to data security, algorithm transparency, regulatory compliances and ethical implications. These factors are pivotal for fostering trust, mitigating risks and ensuring responsible AI integration within organizational frameworks.
• Importance of businesses maintaining the authenticity and data quality of data models: Implementing rigorous data governance frameworks ensures standardized definitions, improving consistency and trustworthiness across data models. Secondly, continuous validation against real-world outcomes further guarantees relevance and reliability over a period of time. Also, leveraging advanced analytics and AI-driven techniques promotes detection of anomalies early on, preserving the accuracy further. Finally, fostering a culture of data stewardship among teams promotes the cultivation of a proactive approach towards data quality maintenance, ensuring ongoing refinement and adaptation to evolving business needs.
• Implementing AI (and the associated implications) in case of poor data quality: Implementing AI with poor data quality requires rigorous evaluation of data cleansing methods, implementing robust algorithms resilient to outside noise, interpretability of results amidst all the uncertainties and strategic planning to mitigate risks – all of which would ensure actionable insights and reliable outcomes. Furthermore, actions like the creation of connected systems, precisely defining roles and responsibilities and ascertaining and benchmarking the quality of data that is expected further ensures setting up a level of standardization.

• AI helping businesses demonstrate ROI on their data initiatives: Key factors in this regard include clear problem definition aligned with business goals, KPI mapping for monitoring and understanding the role of AI and for the purpose of robust data quality management. Additionally, sophisticated AI algorithms could be tailored to specific use cases, and an overall structured approach to measuring both quantitative outcomes and qualitative improvements in operational efficiency would further ensure a demonstrated ROI.

The last word

The session successfully presented strong thought leadership towards the common cause of ‘AI enabling improved Master Data Processes’ through a panel consisting of industry leaders having diverse professional backgrounds. It drove home the ultimate point that ‘mindful use and implementation of AI through robust governance frameworks and drilled down training models backed up by high benchmark data quality’ would enable continuously improving and evolving Master Data Processes on an enterprise level.

Author: Apoorva Dawalbhakta, Associate Director (Research) at Quadrant Knowledge Solutions