Speaker Q&A – Julie Meloni – Senior Engineering Manager, Improbable & Former Director of Product & Strategy/Operations, USDS

Can you tell us about your experience during your 2-year tour of duty with USDS? What was your most proud moment?

I gave a talk in 2018 called “The Hardest Job I’ve Ever Loved”, and that’s what my time at USDS was. I had no expectations going into USDS; on my first day, when I had the obligatory chat with the Administrator and Deputy Administrator (at the time, Mikey Dickerson and Haley van Dyck), I asked them exactly what on earth made them want to hire me?  I was, and still am, an absolute nobody. I never worked anywhere fancy for a huge salary and stock options and free food, and I’ve never given a TED talk or raised capital to fund a startup. It’s true that I’ve been working in and around “tech” since the early 1990s, but so have a lot of people. I got so fed up with Silicon Valley culture around the 2nd dot-com boom that I took a break and got a PhD in English before finding my way back to Engineering and Product Management.  As you can see, nothing about my background lines up with the popular notion at the time that “the White House is hiring the best and brightest geeks from Silicon Valley.”

But what Mikey and Haley said on that first day was that my background and my interviews made me sound like a person who has seen a lot of technology, a lot of bad software development and possibly even some good software development, and who has managed to get (figuratively) kicked in the face a bunch of times but kept getting back up and trying again.  They were right; that’s a great way to characterize the previous twenty years, but it was not exactly what I thought I’d hear on the first day at my new job.  A week later I found myself working with the Digital Service team at the Department of Veterans Affairs, just trying to do anything to help move some important work along so that some of the most deserving and underserved people in the country could get the benefits they earned for the service they performed.

That’s how I spent the next two years – just trying to helpfully move work along.  I had some focused time on a few projects, like identity management at the Department of Veterans Affairs and a cross-agency project around H-2A Visa Modernization for temporary agricultural workers, but I spent more time in a leadership position as Director of Product Management and Strategy/Operations.  In that role I was responsible for the well-being and professional growth of around 45 people, as well as project prioritization and people allocation for USDS projects, and evaluating and selecting candidates to hire into our weird cross-functional matrixed term-limited environment.  I’m most proud of the work that went into enabling all the project teams to do what they did best, which was to create user-facing services for the hundreds of millions of people served by the US government.  I’m also really proud of working with a truly bipartisan group of folks to get the outdated wording in a law from the 1980s changed to reflect the way that people communicate in the almost-2020s. This meant that perfectly safe technical solutions were able to replace paper solutions that were only in place because of said law (come to my talk to hear more ridiculousness!)

As a pioneering woman in the USDS, do you think your role inspired other women to take up roles in tech?

It’s kind of you to think that, but I’m not a pioneering woman in the USDS.  I’m just one of a long line of people who were lucky enough to do a tour of duty.  The USDS was a bit over two years old when I joined, and some amazing women with far greater ability to weather bureaucratic storms than I will ever have, were some of the organisation’s founders.  When I think of pioneering women in USDS, I think of Haley van Dyck, Jennifer Pahlka, Erie Meyer, Dana Chisnell, Vivian Graubard, Jennifer Anastasoff, and Marina Martin – and then I feel bad for not naming literally every other woman I worked with at USDS – but there were a lot! Numerous women have served as individual contributors, communities of practice leaders, and directors of agency teams since the founding of USDS.  Every single one of them is an inspiration for someone to think about spending time in civic tech as an engineer, product manager, UX researcher, or bureaucracy buster. However. the inspiration to join USDS (or even tech in general) should be fed by the work.  At USDS, the work is inspiring, and there’s plenty of it to be done.  I hope people join USDS, civic tech, or just tech in general because of the problems that still remain to be solved.

If I am any inspiration at all, I would want it to be for the people with non-traditional backgrounds, who see that you don’t have to go to a specific school and major in a specific field. Nor do you have to do a specific set of internships and work at a particular set of companies in order to make a difference through technology.

Developing soft skills is crucial for the implementation of emerging technology – have you been involved in any programmes designed to build on these and how can the UK public sector learn from this?

My time and work at USDS was far more driven by soft skills than hard tech skills, and I think the vast majority of USDSers past and present could say the same.  Although every project we did at its core had something to do with technology – create a better interface, create a secured system, keep people from dramatically overspending for hardware they didn’t need, and the list goes on – it wasn’t the technology bit that was truly difficult. Rather it was working through layers of human concerns every step of the way.  We were successful when we held true to our organisational core values, especially: “Design with users, not for them” (focus on empathy and curiosity), “Go where the work is” (meeting partners where they’re at, both physically and mentally), “Optimise for results, not optics” (we work for the people, not headlines, and the stakes are high), and my personal favorite “Find the truth. Tell the truth.” (requires humility and the ability to press hard but respectfully, and be confident in the path forward).

The UK public sector can learn from this the same way USDS did at the beginning – by looking closely at what you’ve already written! It’s no secret that USDS was modelled very closely after the GDS, and the original Government Design Principles published in 2012.  Some key soft skill elements of those first principles were “start with user needs,” “this is for everyone,” and “understand context.”  These principles have evolved into the Gov.UK Service Standard, and many of the fourteen points listed require soft skills far more than they do hard tech skills: “understand users and their needs,” “solve a whole problem for users,” and so on.

All of this work requires empathy more than it does the ability to program a computer, but that’s important too.  I don’t know how to teach empathy – there are far smarter and better educated people than me out there to do that – but if you want to introduce new technology and transformative digital services into the government, you must be deeply empathetic to both the customers and the users of that technology.  Without empathy in your product development, you will end up with yet another generic piece of software that covers a limited set of use cases. Without empathy at the discovery, MVP, or implementation point, you are likely to have stakeholders actively working against you in an environment where stasis rather than innovation is rewarded.  Start together with your users from day one and understand their pain points and what success looks like to them, because even the coolest technology doesn’t matter one bit if the people who need to use it, can’t.

How does Improbable’s technological focus help people solve problems, answer questions, and lead to better functioning governments?

The company I work for now, Improbable, looks like a video game platform company at first glance — and it is!  But some parts of the underlying technology we produce for the gaming world also have applicable – and incredibly powerful – uses in the government space as well.  For example, our modeling and simulation capabilities enable the creation of synthetic environments of unprecedented size; these environments can be used for specialized military, disaster-recovery, and other data-intensive training programs and simulations.

We work with our customers in the government to help them better understand their most challenging problems by designing solutions with them that leverage probabilistic techniques (among others) to run and operate simulation, training, and wargaming environments at a scale, fidelity, and complexity that they haven’t been able to achieve before.  You could boil all that down to this: people (and governments) can think more deeply about potential answers to hard questions when they have more data-driven and realistic scenarios to work with, leading ultimately to better preparation and outcomes.

If there were no legislative or budget issues, what does a completely technically-able government look like?

I know I’m probably supposed to answer this question with something like “that would be awesome!” but governments function (for some degree of “function”) because of some amount of legislation and some amount of budgeting, so it’s not a realistic question – I reckon if we lived in anarchy we’d have different problems and a more technically-able group of people would probably make the situation worse.

I’d be perfectly happy, and plenty of us could get on with the work of serving people through reasonably new technology, if TQ (technology quotient) always sits with its friends EQ (emotional quotient) and IQ (intelligence quotient) at the table where laws are hammered out.  That’s it.  Maybe then we can start getting down to the business of letting computers do computer things  — e.g. objective determinations really quickly — and letting people do people things – e.g. those which require subjective determinations but sure can be made easier with significant amounts of objective data – and we’d all get somewhere.

We’re still at the stage of tens of thousands of people doing computer things really slowly because one badly-written clause in a law somewhere says that’s what they have to do.  I can’t think past that, no matter how much I want to be innovative and work toward technological utopia in the government; the reality in the US is that Veterans have died while waiting years for their disability claim to be processed. It had then been found that a required paper form wasn’t filled out by the right person at the right time years earlier, and the required paper notice never made it to the Veteran because the agency has multiple (and no canonical) physical addresses for that person.  In a completely technically-able government, that would not be the normal situation it is today.

You can catch Julie’s talk at the GovTech Conference on 2nd October 2019!