From Program Management to Program Efficiency and Innovation

By Tela G. Mathias

Traditionally, federal housing agencies, regulators, and the mortgage industry as a whole have relied on Program Management Offices (PMOs) to implement new systems and processes. As waterfall has fallen out of favor, and especially as the dreaded Scaled Agile Framework (SAFe) has exploded in popularity, PMOs have been renamed as Agile Management Offices (AMOs). They still have all the same problems, even with the new name.

The best PMOs are integrated, essential parts of high performing teams. They comprise people who really understand “the business” or how to deliver value with technology (or, ideally, both). They pick up all the things that fall down. They break down barriers. They help deliver bad news. They keep the team fed.

But more often than not, they become: The Process Police.

They pester us weekly to give data for: The Integrated Management Schedule (IMS).

They nag us to update: The Weekly Status Report (WSR).

Data calls for the IMS and WSR are always a dead giveaway that our PMO has become The Process Police. Why? Because integrated, essential parts of high performing teams don’t need to make data calls. They just know. Because they are essential. They are integrated. They know what’s going on.

With the rise of generative AI (GenAI), maybe there’s a case to rethink the approach. Instead of a PMO that manages projects, what if instead we had a Program Efficiency and Innovation Office (PEIO) focused on driving efficiency, fostering innovation, and unlocking AI-driven improvements? True production implementation of genAI at scale in mortgage remains spotty. Many lenders, servicers, and vendors are still trying to overcome genAI fear, figure out governance, navigate the completely uncertain regulatory landscape, and understand what it takes to make their data AI ready.

Enter the Program Efficiency and Innovation Office (PEIO)

A Program Efficiency and Innovation Office can tackle these issues head-on. The efficiency arm ensures we run a tight ship – on schedule, at or under budget, and delivering the value proposition. The innovation arm stands up the genAI lab environment, mines for use cases and opportunities for additional efficiency, and runs experiments. Efficiency and innovation work in tandem to meet the core business objectives, optimized using the best modern technology has to offer. The goal is to replace lengthy status meetings and rigid project plans with investment stewardship, outcome focused agile experimentation, and incremental innovation delivery. It’s a difficult balance, especially in federal where contracts simply do not support this kind of approach, but one we think can be achieved.

Intersecting Private and Public Sector with Efficiency and Innovation

One of the goals of the PEIO might be to intersect the best of Silicon Valley with the private sector in partnership with the public sector. Let’s take skilled genAI operators, intersect them with experienced (useful) program managers, apply hard work and hustle, and see what happens. Some things a PEIO would need to consider:

1. How to use GenAI to manage projects better. A PEIO could implement genAI-driven tools to track program efficiency, automate reporting, automate research tasks, and mine for opportunities. GenAI can help draft project documentation, perform aspects of risk assessments, and generate compliance reports, reducing administrative overhead. The PEIO will need to be the huma-in-the-loop in these scenarios, ensuring AI-generated insights are explainable and traceable, which is critical to maintaining transparency in decision-making.
2. Standing up responsible GenAI frameworks. The PEIO should work with other stakeholders to establish the lightest weight, responsible AI governance framework that defines AI usage boundaries, ensures model accountability, and addresses regulatory compliance. Where genAI is used to deliver operational value, the PEIO should pay attention to and report out on AI auditability processes, keep documented decision logs, understand validation criteria, and understand bias mitigation (if applicable based on use case for the program). The PEIO could also serve as (or intersect with) the training team to create the AI aware workforce and grow the necessary skills.
3. Enabling the provision of AI-ready data. A PEIO sees around the corners when it comes to AI ready data. Driving the team to data quality standards, ensuring structured, tagged, and machine readable datasets for AI use. It should be aware of data governance policies that cover data lineage, security, and accessibility for GenAI applications. It should work with cross industry stakeholders on achieving AI ready policy data in housing.
4. Ensuring federal and commercial contracts support agility. The PEIO must work with procurement teams to include flexible AI adoption clauses in contracts, allowing iterative implementation. It should ensure AI compliance requirements are embedded in vendor agreements, covering model transparency, data ownership, and security. The office needs to develop contract monitoring frameworks to assess AI effectiveness and adjust terms as regulations evolve.
5. Introducing GenAI labs and an experiment-first approach. The PEIO must help the team establish controlled environments for testing AI solutions, ensuring alignment with business objectives before full deployment. It should aid in development of structured experimentation protocols, defining success metrics, failure thresholds, and rapid iteration cycles. The PEIO should facilitate collaboration between teams to ensure AI innovation is scalable, secure, and regulatory-compliant.
6. Connectivity to the regulatory environment and legal progress. The PEIO must maintain a system for continuous monitoring of regulatory and legal updates, especially now as the landscape is full of uncertainty and constantly changing. It should integrate AI tools that automatically assess the impact of regulatory changes on ongoing initiatives. The PEIO should coordinate with legal teams to preemptively address potential compliance risks before they disrupt operations.
7. Thoughtful approach to scaling solutions, GenAI or otherwise. The PEIO must work across companies and agencies to develop a structured framework for scaling AI projects, ensuring interoperability with existing systems and minimal operational disruption. It should align teams to implement phased rollouts, continuously measuring AI impact and adjusting strategies based on performance data. The PEIO should ensure workforce readiness by integrating targeted training that is right for each role and appropriate to the phase of the lifecycle of the team.
8. Ensuring stewardship of all financial investments. Of course, the PEIO should continuously monitor program operations to ensure sound financial stewardship. More than ever, the PEIO should have a direct line to budget and program executive leadership to talk openly about opportunities for additional fiscal responsibility, while also ensuring funds are available for experimentation and innovation.

The Call to Action

Federal housing agencies and state and federal regulators should, first and foremost, focus on providing AI ready policy data for the industry. They should then look inward to their own readiness for AI and start to take the necessary procurement steps to create agile and genAI ready contracts. They should consider appointing a PEIO leader to drive this type of culture change and evaluate the performance of this new role based on outcomes on a 30-60-90-day time horizon.

Mortgage lenders, servicers, and vendors may want to consider a shift from a traditional PMO to a PIEO. They may want to inventory and assess current processes and methods with an eye towards opportunities for increased efficiency. In parallel, they should continue their responsible efforts to bring genAI to their organizations. They should urge their federal and state counterparts to join us as we try to move the industry forward with AI-ready policy data.