top of page
hero-bg-08.jpg
Profet AI Insight

Know about the latest news and industry views through us

【Profet AI Crossover Talks Fireside Chat: Dr. Chia-Yen Lee's Inspiration from Zero to Infinity】Roadmap Session Keynotes

Dr. Chia-Yen Lee Analyzes the Data Science Team's Journey, Initiating Comprehensive Digital Transformation Roadmap


Dr. Chia-Yen Lee Analyzes the Data Science Team's Journey, Initiating Comprehensive Digital Transformation Roadmap

In the inaugural Crossover Talks series - Fireside Chat "AI Fireside Chat: Dr. Chia-Yen Lee's Inspiration from Zero to Infinity" hosted by Profet AI on January 24, 2024, Dr. Chia-Yen Lee from the Department of Information Management at National Taiwan University shared his experiences leading a data science team through digital transformation. Amidst the wave of digital transformation, Dr. Lee provided rich insights, highlighting how to address challenges in this digital era, find solutions through interdisciplinary collaboration, and offer more inspiration to participants engaged in digital transformation.


Success Metrics and Core Challenges of Digital Transformation

The discussion of success metrics is a primary focus addressed by Dr. Chia-Yen Lee, who delves into the "Taiwan SME Transformation Status and Needs Survey" to explore different industries' expectations regarding digital transformation. Success metrics go beyond internal efficiency improvements and should also encompass customer recognition, with "examining tangible profits from the customer's perspective" serving as a clear indicator. It emphasizes that the success of digital transformation should be reflected in market growth.


With Forbes reporting a high digital transformation failure rate of up to 84%, attention is drawn to the core challenges of digital transformation. It emphasizes the preliminary issues that companies face during the transformation process, such as:


  1. Lack of linkage between business outcomes and financial results

  2. Organizational lack of transformation awareness hindering the promotion of growth and absorption of new concepts among employees

  3. Agile teams lacking micro-management capabilities

  4. Distraction caused by engineering details and predictive models themselves

  5. Over investment in novel and flashy technologies without considering internal processes

  6. Inability to translate into actionable language, such as the lack of unified data definitions across departments and the level of domain understanding within the transformation team

  7. Whether data collection quality and sensor installation locations affect data collection


These challenges impede the smooth progress of digital transformation. Subsequently, key solutions are also shared.


Risk Management and Cross-Disciplinary Collaboration in Digital Transformation

After addressing the initial challenges of digital transformation, the risks arising during the execution process become a matter of particular concern. Dr. Chia-Yen Lee identifies the primary reasons for digital transformation failures, including lack of control over external suppliers and information transparency, talent loss to competitors, overly slow decision-making processes, and misalignment in development priorities, leading to uneven resource allocation. These risks need to be appropriately managed at different stages of the project, especially regarding how uncertainty in predictive model accuracy affects decision-making and the prioritization of project execution under limited resources.


Dr. Lee emphasized the importance of interdisciplinary collaboration, which goes beyond team members possessing knowledge from different fields. It also requires each member to be proficient in two or more domain languages. Here, language encompasses not only technical jargon but also industry-specific terminology from various domains. Such interdisciplinary training facilitates better communication and collaboration among individuals with diverse professional backgrounds, thereby enhancing the success rate of digital transformation initiatives.


At this stage, the role of consultants as facilitators becomes paramount, suggesting the utilization of successful experiences from third-party vendors to persuade internal company personnel. Additionally, recommendations for responsible individuals include initiating the use of AI tools simply rather than viewing them as replacements for work. External vendors can assist in constructing problem-solving processes, leveraging resources from academia and industry to swiftly diagnose issues and identify benchmark problems through AI, achieving initial results, and then gradually expanding to broader scopes. Furthermore, training provided by academic instructors in algorithm design and similar areas can also play a significant role within enterprises.


Digital Transformation Milestone: Establishing a Nationwide AI Ecosystem

In concluding the discussion, Dr. Lee highlighted the future trajectory of digital transformation. He believes that the role of citizen data scientists or data scientists will become increasingly significant. This demand extends beyond professional data scientists to encompass a foundational sense that every individual within the organization should possess. The advancement of AI necessitates more comprehensive management and monitoring, requiring the establishment of an AI Pipeline within the organization to facilitate smoother AI applications and rapid deployment. This requires the establishment of corresponding infrastructure and communication mechanisms internally. Establishing a more comprehensive digital transformation team.


The advanced goal of enterprise digital transformation, "AI for All," signifies a shift where AI is no longer solely dictated from the top-down but rather propelled by members within the organization autonomously. With foundational data governance capabilities in place, management only needs to provide the necessary data and technology to teams while concurrently constructing applications required for daily work. This enables employees to further utilize no-code/low-code tools to autonomously develop applications without relying on AI or IT teams, thereby accelerating AI deployment and enhancing efficiency.

Comments


bottom of page