IN BRIEF:
• Although most companies that adopt GenAI follow a similar implementation approach and methodology, expending large amounts of time, money and effort, many implementations fail.
• GenAI empowers corporate leaders to envision a new horizon for their organizations, leveraging this rapidly advancing technology well past the bounds of simple gains in productivity.
The significance of managing generative artificial intelligence (GenAI) initiatives is underscored by a white paper from Pactera Technologies, a leading global technology company, which indicated that a substantial 85% of these projects fall short. Forbes corroborates that the majority of GenAI projects do not meet expectations, underscoring a problematic trend in the field.
GenAI projects possess characteristics that differentiate them from standard software development undertakings. Consequently, the strategies employed in overseeing and realizing the potential of GenAI projects demand a tailored approach distinct from conventional software project management. To address this issue and enhance AI project management methodologies, this article will discuss the following fundamental principles designed to refine the management of GenAI-related projects.
ESTABLISHING CLEAR BUSINESS OBJECTIVES AND THE IMPORTANCE OF PLANNING
To fully harness the potential of GenAI, it’s essential to establish specific, measurable, achievable, relevant, and time-bound (SMART) goals for the GenAI solution to achieve. This crucial step involves a deep understanding of the underlying business problem or challenge that the GenAI solution intends to address. It is also vital to consider whether GenAI is the most suitable solution, ensuring that the technology is not simply being used for its own sake.
Identify and rationalize potential use cases for GenAI that are in sync with core business objectives. This involves a process of prioritization — pinpointing which GenAI applications can deliver the highest value in alignment with the strategic direction of the organization. By focusing on areas where GenAI can make a significant impact, businesses can channel their resources more efficiently and create a tailored approach that maximizes its benefits. For instance, a common application of GenAI is in knowledge management, which could provide value across the enterprise.
Understanding the project life cycle is another fundamental aspect of managing and executing a GenAI project successfully. Establish the stages the project will go through, including a comprehensive methodology that covers various phases such as planning, developing, testing, deploying, and monitoring the GenAI solution. While each stage of the project is important, an emphasis on key differences in developing and monitoring traditional and GenAI solutions is important. For example, in developing GenAI solutions, model “training” directly impacts the performance of the solution in production. Likewise, monitoring performance for its accuracy and precision would be continuous throughout the use of the solution.
Selecting the appropriate tools and methodology is equally critical. Whether in terms of data processing software, programming languages, and platforms for deployment, these must be chosen with the aim of enhancing the productivity and effectiveness of the GenAI solution.
UNDERSTANDING DEPENDENCIES AND PREREQUISITES
GenAI solutions depend on a process often referred to as “learning,” which involves feeding them a substantial volume of historical business data. This data acts as the foundation upon which the GenAI model is adjusted and refined, making it crucial that this information is of high quality. The principle of “garbage in, garbage out” is applicable here, as any shortcomings in the data can lead to flawed results. As the system continues to process new data, its effectiveness is influenced by the accuracy, completeness, and overall integrity of the information it receives.
Another key aspect is the existing technological infrastructure and the broader system of the company. The current architecture and its capacities must be evaluated to determine how they might integrate with or support the effective deployment of the GenAI system. This includes considering the capability of current systems to communicate with the GenAI solutions and manage the additional workload. Scalability also cannot be overlooked. While GenAI can be a powerful business enabler, it requires the proper infrastructure to unlock its full potential. For example, GenAI solutions require significant computational power to function properly, thus, a powerful hardware component will accelerate “learning” of complex algorithms.
The implications of GenAI on existing business processes are profound. The adoption of GenAI systems can lead to a complete overhaul of current processes, possibly making some obsolete. This makes it essential to perform a meticulous gap analysis to understand the differences between current state and future state business processes. This helps businesses ensure they can capitalize on the advantages GenAI offers while mitigating any operational disruptions.
CROSS-FUNCTIONAL COLLABORATION
GenAI initiatives will require cross-functional collaboration. A diverse team composition is necessary due to GenAI projects intersecting multiple domains, requiring a holistic understanding of each area to create solutions that are not only technically advanced but also practical and relevant to the business. For example, a GenAI solution includes business process, application, infrastructure, and data components. To be able to design the solution, it will require the business unit to define the business problem, legal unit to provide compliance requirements, IT unit to provide data, infrastructure and other system requirements, HR unit to manage change, and senior leaders to drive its adoption.
Adequate training will be crucial in ensuring that each team member can contribute effectively and understand the complexities of the tasks at hand. A data scientist, for example, must understand not just the intricacies of algorithms and model-building but also the business problems the technology is meant to solve. It is also imperative to involve cross-functional teams from the earliest stages. Collaboration should be established from the beginning, mixing technical expertise with business insights and ethical considerations. This allows every aspect of the project to be scrutinized from multiple perspectives, fostering an environment where technical feasibility, business viability, and ethical implications are all weighed and balanced.
This blended approach ensures that the solutions developed are realistic, beneficial for the business, and designed with a consideration of their impact on stakeholders and society at large.
CHANGE MANAGEMENT
One common issue in implementing a GenAI solution is resistance. While people may be hesitant to adopt new technologies in favor of established routines, it’s essential for companies to anticipate this resistance and prepare with strategies to address concerns and ease the transition for everyone involved.
To facilitate adaptation, the company should provide substantial training and dedicated support. Instructional programs designed to enhance understanding of the new GenAI system can empower employees. Additionally, a hypercare support system, which offers intensive post-implementation assistance, ensures that immediate help is available for any issues or questions that may arise during the initial stages of using the new technology.
Stakeholder management is also a critical component in ensuring a smooth transition. Clear and transparent communication regarding sunk costs associated with GenAI systems is necessary, as well as assurances that the investments are calibrated for long-term benefits. Stakeholders must also understand the timeframes involved, from the initial implementation phase to when positive returns can be expected. By managing expectations with clarity, the company can secure sustained commitment and support for GenAI initiatives.
PERFORMANCE MONITORING AND OPTIMIZATION
Determine baseline metrics that act as a standard against which the added value of the GenAI system can be measured. Once the system is operational, the company must assess its performance, leveraging both qualitative and quantitative methods in its evaluation while utilizing appropriate metrics and benchmarks. For instance, the company might compare the output generated by the GenAI system against previously established baseline metrics, such as output produced by humans prior to when the GenAI system was implemented.
In addition to monitoring technical GenAI metrics such as accuracy and precision, the company must measure the impact of the system through a business-focused lens. This means putting a spotlight on how the system influences business metrics, outcomes, and the overall impact on company operations.
REALIZING THE POTENTIAL OF GENAI
The potential of GenAI transcends simple enhancements in organizational efficiency. Its profound ability to generate, model, and interpret intricate data place it at the forefront of driving business innovation.
GenAI empowers corporate leaders to envision a new horizon for their organizations, leveraging this rapidly advancing technology well past the bounds of simple gains in productivity. Through GenAI, businesses are not just improving processes, but revolutionizing their approach to problem-solving and strategic planning, planting the seeds for long-term value.
This article is for general information only and is not a substitute for professional advice where the facts and circumstances warrant. The views and opinions expressed above are those of the author and do not necessarily represent the views of SGV & Co.
Ryan Gilbert K. Chua is the business consulting leader and technology assurance leader of SGV & Co.
