aiSTROM -- A roadmap for developing a successful AI strategy (2107.06071v2)

Abstract: A total of 34% of AI research and development projects fails or are abandoned, according to a recent survey by Rackspace Technology of 1,870 companies. We propose a new strategic framework, aiSTROM, that empowers managers to create a successful AI strategy based on a thorough literature review. This provides a unique and integrated approach that guides managers and lead developers through the various challenges in the implementation process. In the aiSTROM framework, we start by identifying the top n potential projects (typically 3-5). For each of those, seven areas of focus are thoroughly analysed. These areas include creating a data strategy that takes into account unique cross-departmental machine learning data requirements, security, and legal requirements. aiSTROM then guides managers to think about how to put together an interdisciplinary AI implementation team given the scarcity of AI talent. Once an AI team strategy has been established, it needs to be positioned within the organization, either cross-departmental or as a separate division. Other considerations include AI as a service (AIaas), or outsourcing development. Looking at new technologies, we have to consider challenges such as bias, legality of black-box-models, and keeping humans in the loop. Next, like any project, we need value-based key performance indicators (KPIs) to track and validate the progress. Depending on the company's risk-strategy, a SWOT analysis (strengths, weaknesses, opportunities, and threats) can help further classify the shortlisted projects. Finally, we should make sure that our strategy includes continuous education of employees to enable a culture of adoption. This unique and comprehensive framework offers a valuable, literature supported, tool for managers and lead developers.

• The paper introduces the aiSTROM framework to reduce AI project failures through strategic planning, data strategy, and effective team composition.
• It emphasizes detailed data assessments and robust risk management to ensure AI models are built on secure, high-quality data.
• The framework also advises on organizing AI efforts and fostering a culture of continuous learning to drive sustainable innovation.

aiSTROM: A Roadmap for Developing a Successful AI Strategy

The paper "aiSTROM -- A roadmap for developing a successful AI strategy" presents a framework aimed at reducing the failure rate of AI projects. By addressing both strategic and practical aspects of AI implementation, the framework intends to assist managers and developers in navigating the complexities of AI integration into organizations.

Introduction to aiSTROM Framework

The paper highlights that a significant percentage of AI projects fail due to poor strategic planning or lack of expertise. To counter this, aiSTROM provides a structured approach beginning with the identification of potential AI projects. Key focus areas include data strategy, team composition, organizational positioning of AI efforts, risk management, and continuous education to foster a culture receptive to AI.

Identifying AI Opportunities

The aiSTROM framework emphasizes the creation of a list of potential AI projects derived from clear business goals, competitive analysis, and technology trends. The suggested approach involves multi-disciplinary collaboration to maximize the selection of impactful projects. This initial step ensures alignment of AI initiatives with organizational strategy.

Data Strategy Considerations

Data strategy forms a crucial component of aiSTROM, as data is pivotal for training AI models. The paper discusses considerations such as data source availability, legal and security issues, storage options, and strategies to ensure data quality. These elements are critical to address the issues of non-production-ready data and privacy, thus enhancing project success.

Building the AI Team

The scarcity of AI talent necessitates strategic hiring and development of AI teams. The framework suggests diverse skill sets for AI teams, encompassing mathematics, big data engineering, and software development, among others. Hiring strategies such as up-skilling existing employees or acquihiring are proposed to mitigate the talent gap.

Organizing AI Development

The organizational structure for AI focuses on whether to centralize or decentralize AI efforts. A centralized team under a chief AI officer may accelerate organization-wide AI deployment, while decentralized approaches support departmental autonomy. The choice between in-house development, outsourcing, or AI as a Service (AIaas) is presented as a core strategic decision.

Navigating AI Technologies

The framework advises on selecting AI technologies with considerations for explainability versus accuracy, human-in-the-loop mechanisms, and the augmentation versus automation of human tasks. Hardware strategies, such as cloud versus in-house GPU solutions, are also covered, ensuring organizations choose the right tools to support their AI infrastructure.

Establishing KPIs and Risk Management

Setting value-based Key Performance Indicators (KPIs) and managing risks form integral parts of the aiSTROM roadmap. KPIs are crucial for measuring AI project success, while risk management addresses uncertainties inherent in AI projects, such as model biases and security vulnerabilities.

Facilitating a Cultural Shift

Finally, continuous education and cultural adoption are emphasized to ensure sustainability and scalability of AI initiatives. Establishing an AI center of excellence fosters an environment of ongoing learning and integration of AI into the company's core functions.

Conclusion

aiSTROM offers a comprehensive framework for devising and executing AI strategies. It guides organizations through critical decisions regarding project selection, data management, team building, technological adoption, risk assessment, and cultural change. By leveraging this framework, organizations can enhance their prospects for successful AI integration. Further validation in real-world contexts is suggested as the next step in research development.