Program leader · operator · maker of useful things

I make complicated things move.

I’m Sarah Hennessy. For more than 15 years, I’ve worked across government, defense, and technology programs — from classified environments and identity-intelligence work to global cloud programs at AWS. The common thread has been the same: make sense of messy work, get the right people moving, and find a workable path when the obvious one breaks. Outside of work, I have a tendency to do the same thing to whatever problem happens to be in front of me.

The serious work ↓The useful side quests ↘

Sarah, slightly off duty

Availablefor senior program management, delivery & operations roles · selective presentation & storytelling projects

If you’re hiring

Start with the work.Global programs, ambiguity, risk, technical teams, customers, and the occasional impossible constraint.

↘

If you have a messy idea

See how I make it legible.Executive storytelling, decks, structure, narrative, and making the important bit obvious.

↘

If you’re just curious

There are side quests.A pregnancy playbook, a wine-cellar operating system, and a women’s health prototype I built for myself.

↘

Selected stories, not competencies.

I can give you the résumé version. These are the versions I’d rather talk about — because they show what happened when the work got weird.

Details from sensitive programs are intentionally generalized. The point is the judgment, not the confidential architecture.

01 · parking-lot engineering

Plan Bhad four wheels

How do we get to yes?

The demo moved. The objective didn’t.

A customer needed to see a technology demonstration, but the government facility couldn’t accommodate it inside. The goal still mattered, and the standard plan had stopped being useful. So I rented a party van, worked with the technical team to set everything up outside, and moved the demonstration into the parking lot instead. The venue changed. The demo happened.

Constraint

The building couldn’t host the demo.

Move

Rent the van. Move outside. Keep the technical objective intact.

Result

The customer still got what they needed.

02 · vendor rescue

12 → 3months of projected delay

The vendor said a year. I went back to the approved list.

During a data-center build, the secure-door vendor told us delivery would be delayed by a year. The issue sat across security, infrastructure and vendors, which meant it had all the ingredients of an orphan problem. I pulled the original approved-vendor list, called the alternatives, got the right technical people into the room, wrote the escalation and gave leadership a recommendation. They approved the switch.

Constraint

A critical security component with a 12-month quoted delay.

Move

Create a real alternative instead of accepting the forecast.

Result

The doors arrived three months after the original date instead of twelve.

The forecast was not the same thing as the answer.

03 · operating model

15–18leads + SMEs aligned

The customer needed answers. The engineers needed to do their jobs.

On sensitive global programs, customers had a lot of legitimate questions. The problem was that every question could become a direct interruption to a technical team. I didn’t want either side to lose, so I helped build the Integrated Program Office: one clear front door for questions, decisions, risks and escalation — with enough structure to protect the engineers without making the customer feel blocked.

Constraint

High-touch customers + scarce technical attention.

Move

Give the conversation a front door.

Result

Fewer ad-hoc disruptions, cleaner decisions and more room to deliver.

Sometimes protecting the work means creating a better front door.

04 · global delivery

9 × 6workstreams × time zones

Nine workstreams do not become one program by accident.

I wasn’t the engineer designing the cloud. I ran the operating system around the people who were. These were multi-year global programs spanning nine workstreams: infrastructure, networking, security, talent acquisition, data center operations, region services, certificate authority, legal and billing. My job was to make the dependencies visible, get decisions made, surface risk early and keep those workstreams moving as one program.

Scale

50+ stakeholders across six time zones.

Mechanism

Make the dependencies, decisions and owners impossible to miss.

What mattered

People knew what had to happen next — and who needed to make the call.

Sanitized screenshot of a dense AWS calendar day showing meetings spanning morning through late evening

A real Wednesday, sanitized

What “six time zones” looked like on the calendar.

The names and room details are deliberately obscured. The 8:30 a.m. start, overlapping midday meetings and 9:30 p.m. leadership sync are not.

A lot of the job was making the next move obvious.

Evidence I cannot turn it off.

The same instinct shows up outside work: notice a confusing system, make it visible, leave behind something more useful than what I found.

No monetization strategy required.

Useful thing #01 · free guide

I made the guide I wished I’d had.

Being pregnant in Germany as an American meant learning a second system at exactly the moment I had the least spare brain. Doctors, hospital paperwork, midwives, registrations, terminology, translations — none of it was impossible, but all of it was scattered. During maternity leave, I pulled what I had learned into one practical guide for the next woman who would otherwise be piecing it together at 2 a.m.

Business model: none. It seemed useful.

Useful thing #02 · household operations

The cellar got an operating system.

We had enough wine downstairs that memory stopped being a viable inventory-control method. My husband would buy more, we would forget what we already had, and occasionally a bottle would reappear like an archaeological discovery. So I built the cellar a proper system: inventory, quantities remaining, vintages, regions, purchase data, pricing, drink/hold notes, visual flags and filters.

Root cause: arguably my husband.

Sanitized crop of Sarah's wine cellar inventory spreadsheet showing wine rows, stock counts, costs and notes

Useful thing #03 · personal product prototype

Her Season started because I was tired of pretending the inputs were the same.

Most health tools asked for a goal and treated the body underneath it as basically static. Mine wasn’t. I wanted something that understood that Cycling, Pregnancy, Postpartum and Menopause are materially different seasons — and that the same advice should not simply follow a woman through all four.

I built the first useful version around postpartum recovery and used it myself. I kept the daily structure intentionally simple: Fuel, Move and Support. The longer-term idea was to pull in Apple Health/Watch, a smart scale and user-entered context so the system could do more of the thinking for me, not give me another dashboard to obsess over.

Built for myself. Used by me. Never launched publicly.

Actual Her Season pregnancy screen showing trimester, nourishment, progress and third-trimester focus

Actual screenshots from the prototype · built and used by me

Making complicated ideas legible.

I like presentation work for the same reason I like program work: there is usually too much information, not enough hierarchy, and one thing that actually needs a decision.

Portfolio presentation sample showing governance guardrails and a phased endowment plan

Presentation redesign + narrative hierarchy

The information was already there. The yes was buried in it.

The original endowment deck already had the substance and the right information. I did not invent the strategy. I made the things that mattered impossible to miss: reorganized the story, tightened the hierarchy, redesigned the deck and made the ask clear enough that the board could see what it was approving. The redesigned deck got a yes.

Narrative hierarchyInformation designExecutive communicationPresentation redesign

How I tend to work

The part between “no” and done.

Find the real constraint. Make the decision visible. Use the relationships that help work move. And leave something behind that makes the next problem easier.

Find the real constraint.

The status report is often describing a symptom. I want the thing underneath it.

Make the decision visible.

Who owns it? What are the options? What happens if we wait? Put that on one page.

Use relationships as infrastructure.

People answer calls, share bad news early and solve awkward problems for people they trust. I build that trust quickly. Think of it as the golf-course part of program management — minus the golf.

Leave something reusable.

A mechanism, a guide, a dashboard, a cleaner story. The next person should not have to rediscover the whole mess.

Serious about the work. Not especially interested in sounding serious about myself.

I started my career as an executive assistant and program security coordinator inside government programs. I moved through identity-intelligence integration and configuration management, then spent eight years at AWS working on dedicated-cloud programs and eventually leading them globally. The work also took me from the U.S. to London and Stuttgart.

The common thread is not a title. It is that I am comfortable in the space between structure and improvisation — where the rules are real, the stakes matter, and somebody still has to figure out how to get from here to there.

I like people. That matters more than most process diagrams admit. A lot of execution happens because somebody trusts you enough to tell you what is actually going wrong before it becomes the official problem.

15+years across technology + government programs

8years at AWS

5continents supported

300K+historical records integrated on an earlier government program

VirginiaLondonStuttgartPennsylvaniaand a lot of time zones in between