At Celonis’ Public Sector Process Intelligence Day, agency leaders and process professionals came together in Washington, DC for a day of networking and discussion on innovative approaches to modernizing Government. Attendees heard from senior executives and industry leaders who are using AI and Process Intelligence (PI) to improve operations and deliver meaningful value.
During the event, public servants and Public Sector thought leaders sat down with the hosts of The GovNavigators Show podcast and shared their insights on how government processes really run and how AI and PI are making it run better. The following 8 nuggets of the wisdom they shared, you can listen to the full show here or wherever you get your podcasts.
1. Put one leader on the hook
Colin Wardlaw, SVP of public sector at Celonis, put it plainly. Name one executive sponsor, give them a short list of outcomes to own, and measure them against it. Measure what matters: cycle time, backlog age, on-time payment rates, and the number of improper payments avoided. A clear owner and a clear scorecard create the authority to make changes and the credibility to show progress.
2. Build a coalition that can deliver
Aubrey Vaughn, VP of strategy and business development for public sector at Celonis, and Andrew Lieberman, director of partner management at Celonis, noted that no single vendor can unwind decades of custom code. Agencies, integrators, and small firms need a shared playbook that works across mixed environments. Process intelligence makes even a messy stack understandable so leaders can govern the systems they have, not the systems they wish they had.
3. Choose governance over heroics
Tom Harker, former Acting Secretary of the Navy and Under Secretary of Defense (Comptroller)/CFO, U.S. Department of War, offered two rules for financial and enterprise programs. First, build a common platform with standard processes so adoption is possible. Second, turn off the legacy systems. Endless options create drift. Leaders must accept bounded risk, sign the cutover plan, and hold the date. That is governance in practice.
4. Measure what matters to people and to the budget
Sairah Ijaz, managing director for public sector at Celonis and former chief information officer and chief artificial intelligence officer at the U.S. Department of Housing and Urban Development, urged leaders to pair financial discipline with outcomes the public can feel. Track operations and maintenance spending and technical debt, and make sure both go down. Pair those dollars with service metrics such as time to benefit, call center response times, and recertification timeliness. Anchor the scorecard in the statute so oversight stays focused on mission results, not features.
5. Make work visible end to end
A digital twin is not a prettier dashboard. It is a shared picture of where rework repeats, where approvals pile up, and where dollars lose traceability. With that view, program offices, auditors, acquisition, and finance can make tradeoffs using the same facts.
6. Prove it with the agency’s own data
Celonis senior value engineer Carter Levinson described the turning point when teams see their processes mapped with their own data. From there, fixes become specific: reroute exceptions, drain old backlogs, rebalance staffing. Results show up in weekly metrics. Legacy environments often yield the biggest gains because they have been hard to analyze. That is not a reason to wait. It is the reason to start.
7. Where to begin?
Start where trust is built. Improve resident-facing services such as food assistance, cash aid, and permitting. Cutting days from time to benefit eases staff pressure and builds public confidence. At the same time, go after cross-cutting processes such as procurement, grants, audit readiness, case management, and spending transparency. These are measurable and already wired into oversight.
8. Ninety days is enough
In three months, leaders can change the trajectory. Appoint a sponsor. Pick four to six key performance indicators. Inventory approvals and systems. Stand up secure data access and build a first digital twin of one process. Pilot with real data. Separate operational changes from policy or statutory changes. Deliver quick wins and track the results. Lock in gains with controls and service agreements, publish the legacy shutdown plan, and expand to the next process.
Bottom line
Government is not short on technology. It is short on shared facts and the will to turn off the past. Give a sponsor real authority, align partners around the same picture of the work, measure what matters, and retire what holds you back. When you do that, modernization stops being a promise and becomes a practice.