ADVISORY ARTICLE
By Naga Mallika Gunturu
Virginia, USA
The AI market is projected to reach a staggering $407 billion by 2027, experiencing substantial growth from its estimated $86.9 billion revenue in 2022. Successfully managing AI projects is complex and requires a nuanced approach distinct from typical software projects. Here are some of the best practices for effective management of AI projects.
Start with clear objectives
Setting the trajectory of an AI project begins with a pivotal first step: clearly defining the problem you aim to solve. Establish a precise problem statement, objectives, and goals that align with your organization’s strategic vision. This foundational step lays the groundwork for success. Adopting SMART goals—specific, measurable, achievable, relevant, and time-bound—facilitates alignment across teams and ensures focused efforts towards a common goal.
AI projects often entail significant experimentation and require substantial upfront investments. Therefore, it is critical to secure organizational alignment on funding priorities early on. This alignment not only supports financial planning but also reinforces the project’s strategic importance within the organization.
Build a cross-functional team
AI projects necessitate a diverse range of expertise spanning data scientists, domain experts, software engineers, business analysts, and software architects. Additionally, collaboration with ancillary teams such as security, data privacy, legal, and risk management is crucial. Each of these roles and teams plays a pivotal role in different facets of AI project development and deployment.
- Data scientists play a critical role in leveraging data-driven approaches to develop models and drive informed decision-making within AI projects. Their expertise spans data preprocessing, modeling techniques, performance evaluation, deployment, and continuous improvement.
- Domain experts provide contextual knowledge essential for accurate problem framing. They help identify relevant data sources, assess data quality, and ensure that the data used for training AI models is accurate, representative, and aligned with domain-specific nuances and requirements. They also influence the feature selection process to improve model accuracy and relevance and collaborate with data scientists to interpret AI model outputs and validate their relevance and accuracy in real-world scenarios.
- Software engineers design and develop the infrastructure needed to support AI applications, including data pipelines, APIs, and scalable computing architectures. This involves ensuring that the AI models can process data effectively and integrate seamlessly with existing systems.
More…
To read entire article, click here
How to cite this work: Gunturu, N. M. (2024). Leading successful AI projects, advisory, PM World Journal, Vol. XIII, Issue VIII, August. Available online at https://pmworldlibrary.net/wp-content/uploads/2024/08/pmwj144-Aug2024-Gunturu-leading-successful-AI-projects.pdf
About the Author
Naga Mallika Gunturu
Virginia, USA
Naga Mallika Gunturu is an IT leader with 24+ years of experience spanning several facets of delivering mission critical IT solutions – delivery management, project and program management, team management, managing operations of zero down-time services, product management and customer engagement. She has experience across US, Europe and India, working primarily in the banking and telecom domains. She is currently working as a Program Manager managing an enterprise-wide initiative to further AI innovation and to introduce AI capabilities into customer facing products and services.
She is a certified Scrum Product Owner and Agile Scrum Master. She holds a Master’s in Business Administration degree and Bachelor’s degree in Electronics and Communications Engineering. Ms. Gunturu can be contacted at mallikagn@gmail.com
