Meeting the Moment Part 4: The Power of Human-Centered Design in Criminal Justice Reform

In the fourth of a special five-part series, Serena Chang, with special guest Tarra Simmons, discusses how human-centered design could lead to more effective criminal justice programs.
Listen Now
Serena Chang

Product Manager, Recidiviz

Tarra Simmons

Washington State Legislator and Former Inmate

Listen Now


Public policy doesn’t always lend itself to human-centered design because there is pressure to get it right the first time, and there isn’t real-time data to make adjustments once a policy is adopted.

Corrections leaders are like CEOs of major businesses who are not given the data that they need to make decisions. They do the best they can but they just don’t have the tools to make significant changes.

We need to include all voices in the discussion if we’re going to evaluate how the criminal justice system works and what needs to change.

The voice of formerly incarcerated people is rarely heard in public policy discussions, which is why she felt compelled to run for office.

Tarra’s kids motivated her to not give up and go back to crime, but she says it takes hope, support and strength to break the cycle.


IVAN STEGIC: Hey everyone! Welcome to the TEN7 Podcast and the fourth episode of a special 5 part series called “Meeting the Moment: Using Data to Reimagine Criminal Justice

I’m your host Ivan Stegic.

This series is a partnership with Recidiviz, a nonprofit organization that is using data-driven tools to help guide change in the criminal justice system.

Our mission at TEN7 is to “Make Things That Matter,” and given all that is happening in our country right now, this series fits nicely with that goal.

Each episode of this series features an interview with a Recidiviz expert, exploring different issues related to the criminal justice system.

Past episodes have explored why the United States is the global leader in mass incarceration, how government and technology might work more closely together to bring about needed change, and how racism and bias in the data we gather may be a barrier to effective reform.

In this episode, we’re going to take a close look at human-centered design. That’s the idea that putting real human needs at the center of the process brings about solutions that are more effective and easier to implement. How might human-centered design practices help us improve our criminal justice system? And what are the barriers that we need to overcome to get this done?

First, I want to continue with our story of Tarra Simmons, a recently elected state legislator in Washington who spent time in prison and who is using that very experience now to try to reform the system and help others break free of the cycle of incarceration.

We are leading off each episode of this series with Tarra’s voice as a reminder that criminal justice is not about numbers, it’s about human lives. And, it’s about hope.

Here is Tarra Simmons, continuing her story, discussing why she decided to run for office.

TARRA SIMMONS: I really ran for office, because I know that there is no other formerly incarcerated people in the legislature, at least here in Washington state. We still haven’t really found any across the nation either, and I think it’s a missing perspective. When you're creating policies that are trying to improve either behavioral health systems or criminal justice systems. I think I bring a missing perspective. I wanted to make sure that I was able to weigh in on these important policy decisions.

Also, I did it because my representative, who I absolutely love and adore, retired. She served in the seat for 16 years, and she came to this work after being a lobbyist for the defenders. So she always cared about criminal justice reform and standing up for the most vulnerable people in our community. She asked me to carry on her legacy, and I thought, I never really thought about running for office, but I figured if I didn’t do it then whoever would? If you think about the history of America, and how at one time there were no women at all serving as elected officials and then there were no people of color, and then there were no openly LGBTQ people.

Formerly incarcerated people are an identity that needs to have representation also, and we face so many challenges from early childhood through educational systems through healthcare systems, and obviously the criminal justice system. But our voices are never really centered in these discussions about how we’re going to improve systems. I did it for a lot of reasons. I also did it because I love my community here, and I want to make sure everybody has a first chance, so they don’t need a second chance later on in life.

IVAN: Can you tell me about the community that you represent in Washington?

TARRA: I represent the 23rd Legislative District. It’s part of Kitsap County where I grew up and live, and it has pretty diverse perspectives across the county. We care deeply about environmental justice, about healthcare and about social and racial justice. So, I definitely am looking out for a lot of different issues now, not just criminal justice, but also making sure that our environment is really protected. It’s something that’s really important to my community here and making sure that our education systems are providing equitable access to everybody. It’s a beautiful, beautiful place to live, and I love my community and the people in it.

IVAN: You are a Civil Rights attorney at the Public Defender Association, but you’re also the Executive Director of the Civil Survival Project. So, I want to talk a little bit about both of those. Tell me about the Public Defender Association first, and what your role there is.

TARRA: The Public Defender Association, we don’t do any public defense, just so you know it, we are changing the name of the organization. But we do a lot of system work to really get at the root of behavioral health intersections with the criminal legal system. That’s where the Law Enforcement Assisted Diversion Program was first created in 2011, which partners law enforcement and prosecutors with case managers. And, instead of arresting people with low level drug offenses, they actually do divert them on the street to case management and skip the whole court process. And so that was founded there. We have projects within the Public Defender Association, and I am the Director of the Civil Survival project which is a project underneath the Public Defender Association.

It specifically focuses on reentry, and so, my work there is about directing a project that is led by and for, formerly incarcerated people. My staff is a majority of people with experience working on reentry, and so we are doing that through a variety of strategies. We do have a legal aid project where we have lawyers who are assisting clients on their individual cases around some of these reentry barriers, and then we also have a policy manager and a community organizer who are kind of leading our legislative agenda now, because I am a legislator and I’m actually on reduced leave serving. And so they are carrying on the work of the organization in my absence. We do leadership development, grassroots organizing, legislative advocacy and legal services for people who have been justice-involved.

IVAN: We’ll hear more from Tarra in our next episode, but before we begin to talk about human-centered design and criminal justice, I asked Tarra about how she was able to break the cycle in her own life and stay out of prison once she was out.

TARRA: For me, it’s always been about the fact that I have kids that I want to make sure that they have a better chance of breaking these cycles than I do. I’ll tell you straight, the truth, when I was working at a fast food, minimum wage restaurant and they’re garnishing my paycheck to pay these court fines and fees, and I ended up having to go homeless because my house was foreclosed on, and I couldn’t find anybody to rent to me, and I actually had to separate from my kids again while I’m studying for the law school admission test.

During that time I definitely thought, What can I do to make money so I can get a place to live? So I can be with my kids again? I thought, Maybe I could just sell drugs but not use them. Those thoughts alone scared me enough to motivate me to keep going, and I had a good support system in recovery, thankfully, that gave me a place to stay so I didn’t have to actually sleep outside.

I just see if you don’t have something motivating you to keep fighting through all of these barriers, it’s so much easier to just give up and go back to crime and go back to prison. You’ll be driven to that, either through poverty or through lack of hope, lack of connection. So we need to make sure every single person that comes out has hope, connection, emotional support and also enough financial support to survive. And a goal that they’re working towards.

IVAN: Tarra’s story shows just how difficult it can be to break the cycle and escape the criminal justice system. She also reminds us how important it is to have support and hope for a brighter future.

So what do we need to do to provide the necessary support and help people stay out of prison? Maybe the concept of human-centered design could play a role.

To explore this possibility, I’m pleased to welcome Serena Chang, a product manager at Recidiviz.

Serena, welcome. Could you give me a little bit about yourself? Tell me who you are and what you’re doing and what you’ve been responsible for recently.

SERENA CHANG: Thank you, Ivan. I’m a Product Manager right now at Recidiviz, which is a technical non-profit working directly with State partners to help them identify ways to quickly and safely and equitably reduce incarceration. I think I got here, sort of, by accident. I grew up in Missouri where I currently am, just riding out the pandemic in my parent’s basement.

When I was young, I really loved making random things out of household items, like cardboard iPod stands, back when those still existed, or duct tape hats for my friends, and so those things were totally impractical but convinced me that I should probably study engineering. And so I went to the University of California at Berkeley. And it was in college that I really came across human-centered design first, pretty much by accident.

We have this plaza at Berkeley called Sproul where people walk through it to get to all their classes. And every day there’s hundreds of student orgs that hand out flyers and you learn very quickly that you’re supposed to look very stoic or unbotherable when you walk through that plaza, but I did not learn that yet. So, I got handed a bunch of flyers, one of them was for a Human-Centered Design club on campus. And so, I went to that info session, and that’s when I first really got intrigued by human-centered design. The whole premise of the club was working in groups with non-profits and different organizations on problems that they were trying to solve. And that was really the first time that I had encountered anything that was much more open ended, and project-based rather than just an intro to artificial intelligence course.

You are asked to make Pac-Man eat all of the dots in a depth-first search algorithm and there’s really only one right way to do that. But in the human-centered design there was a lot more open-endedness, and it was really just all about telling a story, rather than the specific piece of technology that we were trying to build. And so, because I was always drawn to those open-ended problems and because I ended up being really bad at coding, I became a product manager. So that’s how I ended up where I am today.

IVAN: So, Tarra quite literally talks about user experience. Her user experience she talks about what the human user experience is when you leave prison, and what in her mind should be in the place at the time of coming out of prison. If there’s anything, anywhere, we should be focused on human-centered design, it's probably when we’re reintegrating people into society. So you talked about human-centered design. What is human-centered design?

SERENA: Human-centered design at its core, it’s really simple, it’s all about just understanding who your user is, talking to them a lot and understanding what their needs are. And then iterating towards a solution that actually feels like it meets those needs. The great thing is it’s not really a skill per se, but it’s almost like a process, like a recipe book that you can follow, so you don’t need to be a professional chef in order to follow a recipe and make good tasting food, just like you don’t need to be a user experience designer or have a trained background in design to actually practice human-centered design. In fact, I’m certainly not a designer. And you can ask any designer at Recidiviz, and they will not let me anywhere near their box or their design files. And so, the fact that I’m talking to you now is kind of evidence that you don’t need to be.

IVAN: So, you mentioned a couple of key ingredients in the human-centered design process. One of them is talking to your user, talking to your human, the other one is iterating. And it strikes me as though those things might be hard with people who are incarcerated?

SERENA: Yeah, absolutely. I think it really comes down to showing exactly why it’s important to actually include people who are directly impacted to the system into the process of creating something for them, because they understand the experience better than anyone else. And there’s no way that I can sit here having never been incarcerated and be able to tell you exactly what it’s like, but I can certainly talk to them, and I can certainly involve them as we’re creating prototypes and as we’re trying to roll out products. And they can tell me, This is not going to work, or, This is exactly what I need, and that is why it is so important to bring them into the process.

IVAN: So, when I think of human-centered design or any kind of user experience design, I usually think about it in terms of the online context, right? Here’s a thing that someone wants to buy on an ecommerce site. How do I create the flow so that the thing gets into the cart and gets into the checkout?

SERENA: Gotta get into the checkout!

IVAN: Gotta get into the checkout! But that’s not the only place where you can apply HCD as we call it. Tell me more about that.

SERENA: So, I think there are those really classic examples of shopping and trying to make sure that people actually finish buying something or something. Like a TurboTax where previously lots of people weren’t able to do their taxes, but TurboTax created something that really gave people the ability to do their taxes on their own. But you’re right, it’s totally not limited to being online. And, I think in fact, some of the non-technical examples are perhaps my favorite.

So I remember in my interview to be an Associate Product Manager at Google, they asked me to critique an object that was in the room. And there was a box of tissues in the room, and I went on about it for like 15 minutes. So I will spare you most of that today, but at the core of it, you think about a tissue box, and the tissues they exist for people when they’re at their very worst states, when you’re crying, when you’re sick and have to blow your nose, when you have to pick up a dead spider. And so when people reach for a tissue, there’s usually something wrong when they reach for a tissue. And the way that the pop-up tissue box is designed makes it so seamless to actually get a tissue, that you don’t have to think about it. And when you take one out the next one is already there waiting for you. I think that’s a very conscious design decision that makes all our lives a lot easier in those particular moments.

One of my other favorite examples is that experiences also involve human-centered design, and so I remember going to the World of Coca-Cola in Atlanta. And I loved the experience there, because they found such a simple way to pass the time while waiting in line. I don’t even remember the actual exhibits, I just remember the experience of waiting in line where the lines were really long and winding waiting for one of the exhibits, and a staff member simply just tossed a beach ball into the crowd and people started passing that beachball around.

So suddenly the line turned into this really fun game where everyone was participating together to keep the beach ball bouncing around and they totally forgot that they were in line. I think the best part to that experience is that in order to drive that behavior change from people being annoyed and tired to totally energized and happy, that probably cost the World of Coca-Cola 99 cents at the dollar store. So, there can be really simple things. And it’s embedded in basically anything that happens, because it’s human-centered design, so anything that involves humans, it’s possible.

IVAN: You very clearly gave some wonderful examples of when human-centered design worked well, when they’re not even online. What about when we don’t use human-centered design, when things flop?

SERENA: Yeah, tons of things flop, all the time. I think when you don’t use human-centered design, you can waste people's time. But you can also end up making peoples’ lives worse. So a simple more benign example is like a microwave, where everyone is just trying to figure out how to heat your food properly. But if you’ve thought about it, it’s actually really hard to know exactly what buttons to press to heat your food properly, and you probably get it wrong a lot. And you have to reheat or try again or it’s too hot. And like what does the popcorn setting even mean on a microwave, and why are popcorn settings on different microwaves so different? Doesn’t that defeat the purpose of having a preset setting to begin with?

So, I think that’s one example of just a simple example of where human-centered design is not necessarily optimized. I think there’s a more dangerous example where I read a study recently that showed how female drivers are 47 percent more likely to be seriously injured in a car crash, because the crash test dummies they used when testing airbags in the crash test mechanisms in cars were using dummies the size of an average male. And that’s what happens when you don’t actually include all of the right people into the design process.

IVAN: Now you’re definitely skewed towards one use case and completely ignoring everybody else. In your experience, does the work that you’ve seen and worked on with Recidiviz and Google and others, does government do any human-centered design?

SERENA: Definitely. Human-centered design in government can happen at many scales. So, some examples that are not necessarily Recidiviz-specific are just thinking about the immigration process in different countries. I have a friend who recently moved to Canada and she posted on Facebook about her immigration experience and how Canada had noticed that she hadn’t used their newcomer settlement services. And they asked in a survey questions about their language fluency, their sense of belonging and why she didn’t use the newcomer settlement services, which I think was a great example of how Canada actually was trying to talk to the people who are using their services, or not as a way to actually design the services and improve them.

And so, more generally you can really notice the way that different countries do customs, and also how streamlined their public services are. So as a driving traveler in a location you can feel totally calm or totally frenzied based on the place that you’re arriving in. So another example is, I lived in Sydney, Australia when I worked at Google for about a year and a half. And the public transportation system is incredible because everything from buses to trains to trams to ferries, it’s all connected to a single payment system and so this makes it way easier for people to use the public transportation which is better for the government and it’s better for the people, and so, I think that’s an example of government using human-centered design to try to meet people's needs and improve it.

IVAN: One of the things that Tarra talked about is wanting to make sure that everyone has a first chance, so they don’t need a second chance later on in life. To me that means that she’s bringing a certain perspective to that human-centered design, she really wants to represent those people in what should be a human-centered design process and perhaps where voices have been missing in the past. What makes it challenging to do human-centered design, with government, in this context? You’ve talked about successful things that other governments have done, but where are the challenges?

SERENA: Human-centered design, I think, almost never gets it right on the first try. That’s the whole power of iteration. But in government there’s a lot more pressure to actually get it right on the first try, because you’re affecting people's lives, and it’s a lot harder to iterate on something once it’s in people’s hands. And so, if you think about the policymaking process for example, where all of the iteration has to happen upfront, there’s lots of changes and discussion that happens between when a policy gets introduced and when it actually gets passed.

But once a law is passed it’s really hard to make changes because there aren’t very many systems in place right now to understand how the policy is being implemented and what impact it’s actually having. And so, I think that’s why it’s challenging and why it’s also important to have real-time data so that you don’t have to wait three years or five years to know if something is working. And that’s why a lot of our efforts at Recidiviz are focused on trying to find ways to help the government actually shorten that feedback loop and make it a little bit more real-time, so that it is possible to iterate even after a change is made.

IVAN: So, what you’re describing is the inability or the difficulty for government to do fast-based changes that are user driven, and so the government is slow.

I know that government’s one of the biggest consumers, if not the biggest consumer of computers and majorly deployed software. I read a stat here that says about 40 percent of users that you have at Recidiviz are still using Internal Explorer 11. Why is that a problem?

SERENA: It’s less the problem that the government moves very slowly. I think lots of things actually move very slowly. So, for instance, at Google when you work on Google Search, there are often 20 PMs that work on search and they work on the same web UI and all of them need to approve a change before it gets out the door. And so, it ends up slowing things down. When you work in government you also work in a very interconnected web.

In criminal justice for example, you work in a very interconnected web of corrections, courts, policing, probation, parole, all of whom are responsible for a different part of it, in trying to move it in a given direction, and so I think the interconnectedness of it makes it naturally very challenging.

And it goes back to what I was saying earlier about it being very high stakes, because it’s important because everything that the government does impacts everybody. So, if you work in tech, if you work on a gaming app, you are only impacting the people who care about that particular game and have downloaded your app. But governments and criminal justice, it’s not an opt-in system like that. It’s a very different kind of responsibility and that’s why it needs to be taken very seriously and often why it’s harder to iterate once something gets into people’s hands.

IVAN: So, besides iteration, what are the biggest challenges facing corrections officials today?

SERENA: Something that people don’t think too much about is that corrections is an 80 billion dollar a year industry, and that each state corrections leader is fundamentally basically one of 50 people running it. They have hundreds of staff and thousands of people under their care, and so, any CEO of any company that has that type of scope would have real-time profit and loss metrics, they would have dashboards that tell them how many people are using the product that they just launched.

But corrections officials often don’t have that feedback loop to tell them how it’s going. They don’t have the data in real-time which makes it really hard, especially when unexpected things like COVID happen. And there’s no quick way to help them decide what to do and understand exactly how their actions are playing out. So, I think that’s one of the pieces that makes it a lot harder is that there is limited data and not a real-time feedback loop to make those decisions.

IVAN: So, what we’ve talked about in previous episodes is fragmentation and fragmentation across not just corrections systems, but across courts and the parole system and all of these things that are separate but how you’re sort of trying to do the same thing. How does that affect the Recidiviz product and any product that’s trying to address these holes of not having real-time data for leaders to be able to react to?

SERENA: It’s tough because corrections sits at the backend of such an interconnected web as you were saying. I was trying to think of a good analogy for this, and it’s almost like if you’re running the TEN7 podcast, but no one at TEN7 actually gets to choose who comes onto your podcast or what they talk about. That would be very hard, and that’s often the challenge that corrections officials face.

So, you’re right, it is hard because there are so many people involved. But I think the basic concept of iteration in human-centered design isn’t that different than working in any other context, because we have to start out by getting the 360 degree view of the problem by talking to people who are impacted by the system, people who are running the system and the people around the system. Like advocates and lawmakers who are trying to move it in a given direction.

And so, when you understand people and know what their needs are, what perspectives they’re bringing to the table, it becomes really clear that many people want different things. So, for instance, a criminal justice researcher might want access to just the raw data so they can do all of their own analysis. And a lawmaker might just want a one line summary of the takeaway on an 8½ x 11 piece of paper so that they can write notes on it and read off of it in a legislative session. And if we try to build a generic web dashboard neither of those people will be happy. So, it really starts by understanding all of the people and the unique perspectives they’re bringing into the room.

IVAN: And all of these unique perspectives are also from different parts of the process, aren’t they?


IVAN: So, corrections sits at what you call the back half of the criminal justice system, and they don’t control the people that are actually being sent to prison and that are being paroled, right? So how do you connect across these difficult layers?

SERENA: One piece of it is making sure that we are involving all of the right people in conversations throughout the entire process. So, if we’re building a dashboard for corrections, we are not just talking to the corrections leaders, we’re talking to their staff, we’re talking to the people on parole and probation. We’re talking to the legislators who are making the policies that impact the numbers that are showing up on the dashboard. So being able to talk to all of them repeatedly and as we’re building out each step of the process, helps us understand what is the thing that we can help corrections take action on that we can be building.

IVAN: Are you talking to prisoners as well?

SERENA: We talk to people on supervision basically everyday and understand what their experience is like having been in prison, being on parole or probation right now. And what their unique experience is.

IVAN: And you use that data to influence and to inform the design of your product?

SERENA: Yeah, absolutely. We’ve learned a lot of things that we never would’ve known if we had not talked to clients on supervision. So, we’re often trying to build tools for the parole and probation officers. And talking to the clients helps us know, Are these things actually going to work or will they create these unintended impacts? So, one client that we were talking to said, The best thing that my parole officer could do for me is just acknowledge that I’m a human, not just a number and a crime.

That impacts a lot of design decisions, like, we probably should be surfacing the client's name and not their department of corrections ID or number. There are lots of simple things like that, that we can completely overlook if we are not actually talking to the people impacted by the system.

IVAN: Tarra mentioned an anecdote that was related to the Public Defender Association. Instead of arresting people with low level drug offenses, they were actually doing things to divert them on the street to case management to skip the whole court process. And to me that sounds like an organization and a system that has thought about what it’s like dealing with low level drug offenders and have used a human-centered design process to change the way that those are dealt with.

SERENA: I think we prevent unintended consequences by really being open to challenging our assumptions at every step of the process. And so, when we set out to design a new tool, we first figure out what change we’re actually trying to drive, so let’s say that change is to increase the number of people who get off of supervision successfully. But then we think about all of the things that can go wrong and involve a lot of people into that process. And so, maybe for example, increasing releases from prison actually increases the racial disparities in prison.

We make sure to identify those things that we do not want to happen, and we track those in an automated way day over day. And as the products roll out, if any of those metrics that we’re paying attention to, processor and threshold, then we talk to the people using our product about what’s actually happening, and we find ways to address it, which might involve pausing or pivoting or trying something new.

And so I think it really all comes back again to iteration, and that we will not avoid unintended consequences if we think we have all of the answers at the beginning. But being able to iterate quickly and challenge our assumptions is what will help us prevent them long-term.

IVAN: How receptive are the people you work with to this kind of iterative process? Especially when such important consequences are possible?

SERENA: We’ve been excited by how receptive people are because of a couple of things. One is that it is a very small investment upfront, and we start with something small. And so the risks are a little bit lower when we start with a very small group of testers, or a really small piece of technology. Versus if we started with a 10-year project that had a giant system migration involved that would impact operations 100 percent across the board. That’s a much larger risk that I think people would be less willing to take. So, starting with something small and really affirming that we want to partner closely with them to be able to iterate on that small thing to make it even better moving forward is something that I think people tend to be pretty receptive of.

IVAN: Yeah, lower risk seems to make it easier to bring you on board. What’s an example of something that would be a small low risk thing that you were involved in?

SERENA: One example of something where we start small is if we’re building a set of tools for parole officers to help them manage caseloads and improve outcomes for people on their caseload, we won’t start by launching the tool to all 1,000 parole officers in a given state. We might start with six parole officers who we identify as trusted testers. And we can actually talk to every one of them to set context that this is the very first thing, this is the smallest thing that we will launch, we want your feedback in telling us if this is not useful or if this is actually very confusing or actively harmful.

If we start with six, we have much more control over how we can iterate, and we can bring them into the process to help us improve that product so that by the time we roll out to 30 we will feel like we’ve de-risked ourselves a little bit. And then we can learn from that group of 30 so that by the time we roll out to 100 we will feel more confident. And we’ll learn more things from that 100, and then by the time we roll out to thousands of people we will have already had these gating steps where we make sure that all of the metrics are pointing in the right direction, and that we’re not driving any behaviors that we don’t intend to be.

IVAN: You mentioned earlier that there are some unintended consequences as well as designs that you might make, and changes that you might introduce, and how you try to avoid those unintended consequences. When you focus on human-centered design in the criminal justice system, what do you think are corrections officials biggest fears?

SERENA: I think one of the things that feels very unique and hard about the job of corrections is that they are responsible for public safety first of all, which is a very big task. And moreover, when bad news happens, everyone hears about it. And when good news happens, they don’t really get much praise. And so if your job is to protect public safety, but you get all of the slander and none of the praise, that feels pretty bad and is something that is likely on corrections officials minds frequently.

IVAN: What are you guys doing to help that situation?

SERENA: We want to reorient the idea that data will only be used to reveal bad things about corrections, and only be used when something is going horribly wrong and to reveal that. We want to use data instead to surface opportunities and find things that actually are worth celebrating. So, for instance, here’s a parole officer who is excellent at making sure people with substance use needs are matched to substance use treatment. Or, here are a group of officers that are really good at helping people find stable housing. We want to be able to celebrate those successes and highlight those narratives rather than just revealing the bad ones.

I think we’ve already seen it happen in a few ways. So, when North Dakota released their racial disparity data publicly, advocates in the local press celebrated them for making that data public. I think there’s a lot of potential for more narratives like that where things can be celebrated rather than critiqued, and I think it’s clear that the public has a role to play in that, in helping corrections move in a direction that they want and create a culture that is more conducive to improving outcomes.

IVAN: What makes your approach unique at Recidiviz?

SERENA: One of the things that makes our approach unique at Recidiviz is the way we partner and collaborate with the corrections leadership and folks within the government. So, government is used to signing these huge 10-year contracts with tech vendors where it takes them 10 years to build a giant system with 1,000 features. Because they don’t know what they want, and they don’t know what they’ll need, but they don’t want to risk being without them. And so, there’s these hugely scoped projects that take forever to build, and once they’re built, the vendors sort of walk away. There are no more changes.

And it’s impossible to know if, let’s say only 45 of those 1,000 features were actually useful, there’s nothing they can really do about that. And so, at Recidiviz we very clearly signal that we don’t work like other tech vendors, and that the first product we launch will come very quickly, and it will be very small, and it will likely be the very worst version of the product that they will ever see. Because we do want to partner with them and incorporate their feedback to iterate and make it better.

And so, we’re pretty excited that the government agencies we’ve been working with so far have seemed receptive about it, because it is a huge investment to pay millions of dollars upfront for the 10-year contract and the 1,000 features. And by partnering with us, we might learn very quickly within the year, that they only need 45 of those 1,000 features to really move the needle. And in fact, that there’s another 20 that weren’t in the original set of 1,000 that they could really use. And so much more quickly we can just build out the set of things they need and iterate towards getting them something better.

So, I think the key that we really try to lean into is getting prototypes and tools into the hands of our users as early as possible and build them into each step of the development process so that we can very quickly iterate towards the thing that actually does work.

IVAN: One of the things I read in prepping for this interview was, Hit the ceiling and raise the ceiling. Repeat. And I think that came from you, and I think it’s such a good way to describe Recidiviz.

SERENA: I think there's a lot of potential to keep hitting the ceiling and raising the ceiling when it comes to working with the government, because each agency has a limited amount of things they might be able to directly control which is the ceiling that they’re trying to hit. But they’re in that interconnected web with all these other agencies who can lift the ceiling for them.

So for instance, corrections might not be able to directly influence the people that are coming into the system through the courts, and through probation, and through the sheriffs. And so they have a ceiling they can reach. However, the legislature may be able to change a policy or a law that actually changes the people who come into the system, and so that’s the example of raising the ceiling.

And if so, we can get into the process of helping corrections officials actually hit the ceiling with the metrics and with the real-time feedback loop of how they’re doing. And then work with the agencies around them to move the ceiling and do that in a repeated, iterative way. That’s how we can, sort of make dents into this problem.

IVAN: So, one of the things I think that has been kind of a roadblock has been procurement and this 10 year project that costs 10 million dollars. That’s a big thing, and you maybe don’t even get what you really need at the end of that project. And I know that government is trying hard to get around those things with 18F and the United States Digital Service.

Another thing that I think that might be a roadblock is how do we scale the work that you’re doing fast. Because you guys are working with state corrections departments and local systems, and this sort of has to be a federal thing right? It has to be done across the nation as quickly as possible. And Recidiviz is doing wonderful work, but do you have any ideas how we can scale this?

SERENA: Got to think about that one, it’s tricky. [laughing]

IVAN: Yeah. [laughing] Because you can’t throw more people at the problem to fix it right? Recidiviz can grow 10X and would you get 10X more people implemented?

SERENA: There are a couple of things that we think about when it comes to scale. I think you’re totally right that it is very hard to scale when there are so many distinct agencies that we’re working with that have so many unique problems. We can’t end up building one product for each state, or each jurisdiction, or each federal level, because we’ll end up with 50 to 17,000 different products, and that’s not sustainable. We are able to start with a very small kernel that does tie many different corrections agencies together.

So, for instance, a lot of corrections agencies across the board worry about recidivism, where they’re trying to unanimously agree that they’re trying to prevent people who leave the system from coming back in, that’s kind of across the board, something that people are trying to drive towards.

We look for the problems that are common to different agencies and start with those. And we learned that by talking to people in different jurisdictions, in different states, which is something that we can do without actually having signed data sharing agreements and deep partnerships with other states.

It costs nothing to just go and talk to an advocate in Maryland, or an advocate in Florida. And so before we even try to scale the technologies, there’s a lot, we can learn just from talking to people and understanding the common themes that are important to them and solving for those first.

IVAN: Do you think open data and open technology has a role in it?

SERENA: Definitely. I think there are lots of people who are working in the criminal justice space, and one tech company cannot and should not solve all of the problems. But there are lots of researchers and academics and technical assistance providers that are doing great work. And so being able to make the data open source and available for other people working in the space to build upon it and to build their own analyses and to build their own tools, will magnify the impact that everyone can have.

IVAN: Makes sense to focus on the human and the design to be centered around the human for the criminal justice system. It absolutely makes sense to do that. What are you hopeful about for the future and why do you keep working on this?

SERENA: I’m hopeful because every day I talk to new people on supervision, talk to new advocates, corrections leaders and lawmakers who are all looking at the problem from a very different angle, and they’re all inspiring to talk to. I don’t think that tech can or should solve all of the problems; it is one very small piece of it. And I do think that human-centered design is a tool and process that can bring a lot of those people together, and unite a lot of the people who are working to solve this problem.

And so, knowing that there’s a very small thing that I can do and a very large set of things that I can learn from people also working in the space is what really keeps me hopeful. I was hanging on the Sydney beaches for a year and a half before starting this role, so I feel like I have a lot of juice left in me to keep working on this for a while.

IVAN: [laughing] Sydney beaches, they sound pretty amazing right now. Thank you so much for spending your time with me today, Serena, it’s been great talking to you.

SERENA: Absolutely. Thank you as well. This was fun.

IVAN: Our series will conclude next week with more from Tarra Simmons, and a discussion with Terin Patel-Wilson, a product manager at Recidiviz.

Terin will help us look for signs of hope and examples of criminal justice reform efforts that indicate that we might finally be ready to tackle this issue in a more comprehensive way, and perhaps stop being the global leader in incarceration.

Here’s some of what Terin had to say:

TERIN PATEL-WILSON: Folks are talking about mass incarceration in a way that is exciting. A lot of it has come from advocates. A lot of it has come from things such as social movements like the Black Lives Matter movement, which has made sure that all of our, as we talk about criminal justice reform, we’re doing it in a way that is centered on racial equity.

Some of the things that are hopeful are a lot of the reforms that are being talked about not only reduce incarceration, but are also doing it in a way that attempts to address some of these inequities. So like, differences in sentencing based on drugs. We’ve had recent pushes for the legalization of marijuana and also expungement of the records for folks that were previously arrested. Some elimination of mandatory minimum prison sentences, which have impact in different states in different ways but generally have impacted communities of color more.

Having that being front and center obviously is a harsh mirror to see, but is exciting that we’re looking at the problem head on.

IVAN: Join us next time for the final episode of our series, Meeting the Moment: Using Data to Reimagine Criminal Justice. We hope you’ll subscribe. You can find out more online at Thank you for listening.


This is Episode 117 of The TEN7 Podcast. It was recorded on March 29, 2021 and first published on May 5, 2021. Podcast length is 48 minutes. Transcription by Roxanne Chumacas. Summary, highlights and editing by Brian Lucas. Music by Lexfunk. Produced by Jonathan Freed.

Please rate our podcast! Doing so helps spread the word about the show. Just pull it up in the Podcasts app and scroll all the way down, hit the stars and you're done! Thank you.

Continue Reading

Subscribe and listen wherever you get your podcasts.