CANA's Approach to Generative AI and Predictive Modeling in Project Management

By Jack Murray, Senior Project Manager

Generative AI and predictive modeling are rapidly transforming project management workflows. Every organization in the federal space is facing a significant hurdle: the need for trust-based AI. This is where we lead. At CANA, we are not just observing this transformation; we are actively engaging with it to deliver superior results for our clients and our own organization. Our domain experts and subject matter experts (SMEs) are the validators, verifiers, and accreditors: they drive trust through human-centric approaches to an AI transformation. Our CANA AI strategy for the Project Management practice focuses on three key areas: collaboration, implementation, and validation. 

New AI tools give the organization the speed to deliver quickly, but our domain experts instill trust through collaborative stakeholder development. By maintaining this balance, CANA ensures our program and project management team remains as agile as the technology we use and as reliable as the professionals who lead it.

Working with AI

At CANA, collaborative work is driven through project leadership. We believe the future of project management lies in a symbiotic relationship between the domain expert and the latest Generative AI capabilities. We are actively working with the latest AI models to enhance decision-making and automate routine tasks. Our team is focused on identifying high-leverage areas where AI can truly augment human capacity, such as project kickoff and closeout, predictive risk assessment, resource optimization, and automated reporting. This evolution is made possible by the close partnership between our technical enablers—such as our data science team and systems engineers—and our domain-expert project managers, ensuring that the AI tools we build are both technically sophisticated and deeply rooted in real-world project requirements.

This collaborative approach ensures that AI serves as a powerful tool for our project managers, allowing them to focus on strategic oversight and client engagement. CANA knows that leadership drives excellence, and tools such as generative AI and predictive modeling can drive more efficient day-to-day work. We understand the balance, and we achieve more through transformation.

Implementing AI

Implementing AI into project management is more than just adopting a new tool—it’s about integrating intelligent processes throughout the project lifecycle. We believe that for Generative AI and modeled results to be trusted, they must be understood by project stakeholders. This was the driving force behind a recent session at our CANA Program and Project Management (PPM) Offsite this past fall. Instead of a theoretical lecture, the session focused on AI in our daily routines. We gathered our project managers and Generative AI SMEs to collaborate in integrating Generative AI and prediction into their workflows.

As a result, we are implementing AI-driven tools that can parse vast amounts of project data to pull key performance indicators (KPIs), identify patterns, and flag potential issues far earlier and faster than traditional methods. These tools assist in:

• Client & Executive Level Reporting: Generating and automatically updating deliverables status, financial, and risk management reports.

• Schedule Optimization: Generating and adjusting project timelines based on real-time data and historical performance.

• Meeting Intelligence & Action Tracking: Utilizing AI-powered transcription and summarization tools to capture high-level strategic discussions. 

• Workload Balancing: Distributing tasks efficiently across teams to prevent bottlenecks.

This is where CANA’s domain expert provides the final human in the loop check, be it Project Manager or a Technical Delivery Expert, the up-front collaborative build of a system, or that final verification ensures that nuanced technical details are captured correctly before action is taken or they reach the client.

The most important takeaway from our PPM offsite was that AI tools at CANA are designed to augment, not replace, the domain expert. By offloading the repetitive data-heavy tasks, our project managers are freed up to focus on what matters most: strategic relationships and complex problem-solving. This implementation process is guided by our commitment to transparency and ethical AI use, ensuring data privacy and unbiased decision-making throughout our projects. 

Validating AI Use

The effectiveness of any new technology must be rigorously validated. At CANA, we have established clear metrics for validating the use of AI in our projects. This includes comparing AI-assisted project outcomes against similar projects managed without AI, and focusing on key performance indicators (KPIs) such as accuracy, credibility, on-time delivery, budget adherence, and client satisfaction.

We are continuously running controlled studies and collecting feedback to refine our AI models and integration methods. For example, this article was supported by AI that parsed our internal AI project management practices and values, but has been revised and validated by our project managers. Innovation in AI models is crucial in today’s rapidly evolving world, but ensuring their accuracy and efficiency is the key. Our validation process is designed to ensure that every implementation directly translates into tangible improvements for us and for our clients—smarter planning, more efficient execution, and more successful project outcomes.

Our goal is not just to use AI, but to be a leader in demonstrating its validated, value-driven application in the complex world of project management.