Podcasts

A tiny nonprofit is using tech to quietly transform the prison system

Recidiviz wants prisons to know what's going on with their inmates. The company's open-source software might actually help them do that.

A tiny nonprofit is using tech to quietly transform the prison system

Recidiviz went from a slow, under-the-radar push to integrate data tools into a few state prison systems to a sudden burst of activity to provide the data state prison administrators needed to launch early release programs.

Image: Recidiviz

The pandemic made success stories out of all kinds of for-profit tech companies — and also for one small technical nonprofit called Recidiviz. The company provides open-source data tracking and management software for state prison, probation and payroll systems, and it became astoundingly important in the middle of the deadly pandemic that ravaged prisons across the country.

Clementine Jacoby's Y Combinator startup went from its slow, under-the-radar push to integrate data tools into a few state systems to a sudden burst of frenzied activity to provide the data state prison administrators needed to launch early release programs. Her team helped hundreds of prisons get tens of thousands out of the system on early release in the last year, and now they're hoping they can help some of their clients get a handle on all of the data about what happened, why early release worked and how it could be used in the future.

Protocol sat down with Jacoby to talk not just about the last year, but what it's like hiring top tech talent for nonprofit pay, why engineers want to work for mission-driven companies and what other future founders might learn from the path she took to become a founder of a nonprofit.

You can listen to our full conversation on this episode of the Source Code podcast. Below are excerpts from our conversation, lightly edited for length and clarity.

Talk me through the origin story of Recidiviz. It's an open-source platform, and it's a nonprofit, but this could easily be a for-profit venture. So walk me through how it came to be in this particular form, and why.

Yes, we actually went through Y Combinator. And so we were told many times per day that Recidiviz could be a for-profit. The short answer is that we always knew we wanted to do this as a nonprofit. Nobody at Recidiviz is here to make money. I think everyone here essentially took big pay cuts to do this work.

I think the important thing for us is that being a nonprofit allows us to be opinionated about what we build, and to not need to sort of chase opportunities that are lucrative, but that won't, in our view, solve an important problem. And that is a very big deal, we needed to have that freedom to have the impact we wanted to have.

I think the argument that's often made is you won't be able to attract the right kind of talent as a nonprofit. And in my experience, that has just not been true. I have never worked on a team that's been this technically strong. And my experience has been that there's almost an overwhelming number of people in tech who want to use their technical skills for good, and that that is the much bigger gap. There are lots of top flight engineers who want to make the world a better place, but they also want to report to top flight engineers, and they want to execute on complex technical problems in places with strong engineering cultures. And finding that balance of factors is actually the hard thing.

So why did you go through Y Combinator? And the perspective you just shared about what people are actually looking for is one that I don't hear very often, so I'm wondering if you can explain that a bit more.

So YC's mission is to help these young companies build something that people actually want. And my background was in product design and user experience research [at Google]. And to me, the biggest challenge in building something important is that there are so many distractions along the way to building something people want. And so the experience of going through YC was really valuable.

Everything about Y Combinator was relevant to what we were trying to do. We were the only nonprofit in the batch, but it didn't feel like a wildly different experience from the one the other companies were having.

There's a lot of skepticism toward technology applied to prison systems in general. How do you deal with that fear and make sure that the technology is not being misused?

The first thing I'll say is that the technology itself is not good or bad. And if it were that simple, it would be all much easier. And I say that less to finger-wag at that framing, but more to emphasize that these dangers are real. We need to get more nuanced in a hurry about how we think about technology. Technology will tend to make things faster and more scalable. The rest is all very thoughtful product design, it's talking to PEOs, it's talking to leadership, it's talking to people on supervision, it's trying to design a system that gets the right information to the right person at the right time.

On the flip side, right, if you inject that kind of horsepower and efficiency into a process that has bad bones, you end up in a bad place much faster than you otherwise would have. So you can scale up bias very quickly, you can scale up surveillance very quickly. And it all happens so fast. Before you can even kind of measure these unintended consequences, you're already there. So we take this stuff borderline obsessively seriously.

Our first employee did her academic research on how criminal justice algorithms could become biased and what to do about it. And so now, every feature that we launch has a set of success metrics, and it also has a set of backstop metrics. So if you're launching a tool that's designed to speed up the number of people succeeding on supervision, you're also going to be measuring whether or not it is unintentionally increasing the number of Black and brown people on supervision as a sort of percentage of the whole.

Are these tools doing what they were intended to do? And also, what are they doing that they were not intended to do? You need to measure both of those things with kind of equal fervor, and you need to do it at very short time intervals. Because again, what technology does is speed things up. And you need to know right away what's happening when you launch not just a new tool, but even a tweak or a new feature.

So over the course of the last year, with the pandemic, what has been the hardest part? The unbelievably challenging, frustrating part?

It did seem like there was a non-zero chance that an airborne pandemic in overcrowded prisons was going to be really bad. And so we sat down over a weekend, and we took a bunch of cruise ship data and data from the 1918 Spanish Flu in San Quentin. And we tried to give folks a tool that would give them a basic sense of how many people were going to get sick. And then we released this tool, and we braced for impact.

Suddenly, the people that we work for and work with were in a really hard position of high-responsibility decisions that needed to be made at high speed. And, you know, the big thing that the data showed was that what would slow the spread was early releases. And that makes sense, right? You had to de-densify these facilities. So we helped them understand how early releases would help and use that to create a conversation within states and across agencies about what needed to be done. We helped the states that we were in use their data to identify people who could be safely released. And then, importantly, we also used data to help track how those folks did in the community. Going back to this idea that if you want leaders to do things, you need to give them the tools to see whether or not those things are working.

On the one hand, everything about the pandemic was hard, and it was a bad time to be in corrections or in prison. On the other hand, we learned so much more about how this system works. A lot has been learned during COVID. And so I think it's a big opportunity to think about how we move forward.

We almost never talk about the people who run prisons. I think when we talk about prisons, we talk about the people who are inside them. But talking about the people who are in charge of them is something I think we shy away from socially. But you work with these people all the time. So what is it that prison administrators want from people in the tech industry? From society as a whole?

I totally agree that it's a question that doesn't really get asked enough. Because we as society do expect a lot from these people, and we do, in my opinion, know very little about what their jobs are. I think they want to know what's working. It's like being the CEO of a company that has thousands of employees, that's launching all kinds of products and having no sense of which of those products are being used. And the stakes are quite high. If you think about what it really means to do that job, the hardest jobs in the world are the ones where you have tons of responsibility and very little control.

I think what they want, at least from us, is to know what's working, and to know it faster. And those are the things that they need to actually drive the results that society is asking them for, and that policymakers are asking them for. So it's been a real frameshift, I would say in understanding what those jobs are, and how to best help.

What would it mean to help them succeed? And how would you go about trying to do that? And how could technology help you with that? And I think once you start asking the question that way, this is classic user-centered design. This is something that tech does really well when we are thoughtful. Once you start asking the question that way, you see lots of opportunity.

What do you say to people who share your philosophy of wanting to do good mission-driven work in tech, while also working with really talented engineers? What should they be thinking about if they want to be on a path like yours?

I will say that I was not a person who thought that I would found something. Me and my co-founders started by trying to be helpful to people in the field doing good work who are smarter than us. So instead of starting out and saying, "We're really clever, I bet we can come up with a good idea," we just went to people who had problems that they were trying to solve, and we tried to accelerate their progress.

And that ended up doing a lot of things. One, we learned a ton about how this work was being done, and we got to avoid lots of things that had already been tried. By the time Recidiviz was spun out as an independent nonprofit, we had a pretty strong point of view on where we could be helpful. It wasn't our idea — it had been informed by academics in the space, researchers in the space, practitioners, advocates, people who were actually doing work to help government agencies. So I think that's one piece that I recommend.

A related piece is just this idea of listening. We have had people like The Council of State Governments help us to really establish a point of view and figure out where technology can play a role, and where it's not needed. And both of those are equally important. You need to have trusted partners who are domain experts. As a technology outfit, you should not go it alone. You should find the people who know what they're talking about and who have a strong point of view because they've been doing this work for 30 years.

My recommendation is basically to trust your technical instincts and know what you do well, but to also know what you do not know. And to go find people who know the things you don't know, and partner with them.

Fintech

Judge Zia Faruqui is trying to teach you crypto, one ‘SNL’ reference at a time

His decisions on major cryptocurrency cases have quoted "The Big Lebowski," "SNL," and "Dr. Strangelove." That’s because he wants you — yes, you — to read them.

The ways Zia Faruqui (right) has weighed on cases that have come before him can give lawyers clues as to what legal frameworks will pass muster.

Photo: Carolyn Van Houten/The Washington Post via Getty Images

“Cryptocurrency and related software analytics tools are ‘The wave of the future, Dude. One hundred percent electronic.’”

That’s not a quote from "The Big Lebowski" — at least, not directly. It’s a quote from a Washington, D.C., district court memorandum opinion on the role cryptocurrency analytics tools can play in government investigations. The author is Magistrate Judge Zia Faruqui.

Keep Reading Show less
Veronica Irwin

Veronica Irwin (@vronirwin) is a San Francisco-based reporter at Protocol covering fintech. Previously she was at the San Francisco Examiner, covering tech from a hyper-local angle. Before that, her byline was featured in SF Weekly, The Nation, Techworker, Ms. Magazine and The Frisc.

The financial technology transformation is driving competition, creating consumer choice, and shaping the future of finance. Hear from seven fintech leaders who are reshaping the future of finance, and join the inaugural Financial Technology Association Fintech Summit to learn more.

Keep Reading Show less
FTA
The Financial Technology Association (FTA) represents industry leaders shaping the future of finance. We champion the power of technology-centered financial services and advocate for the modernization of financial regulation to support inclusion and responsible innovation.
Enterprise

AWS CEO: The cloud isn’t just about technology

As AWS preps for its annual re:Invent conference, Adam Selipsky talks product strategy, support for hybrid environments, and the value of the cloud in uncertain economic times.

Photo: Noah Berger/Getty Images for Amazon Web Services

AWS is gearing up for re:Invent, its annual cloud computing conference where announcements this year are expected to focus on its end-to-end data strategy and delivering new industry-specific services.

It will be the second re:Invent with CEO Adam Selipsky as leader of the industry’s largest cloud provider after his return last year to AWS from data visualization company Tableau Software.

Keep Reading Show less
Donna Goodison

Donna Goodison (@dgoodison) is Protocol's senior reporter focusing on enterprise infrastructure technology, from the 'Big 3' cloud computing providers to data centers. She previously covered the public cloud at CRN after 15 years as a business reporter for the Boston Herald. Based in Massachusetts, she also has worked as a Boston Globe freelancer, business reporter at the Boston Business Journal and real estate reporter at Banker & Tradesman after toiling at weekly newspapers.

Image: Protocol

We launched Protocol in February 2020 to cover the evolving power center of tech. It is with deep sadness that just under three years later, we are winding down the publication.

As of today, we will not publish any more stories. All of our newsletters, apart from our flagship, Source Code, will no longer be sent. Source Code will be published and sent for the next few weeks, but it will also close down in December.

Keep Reading Show less
Bennett Richardson

Bennett Richardson ( @bennettrich) is the president of Protocol. Prior to joining Protocol in 2019, Bennett was executive director of global strategic partnerships at POLITICO, where he led strategic growth efforts including POLITICO's European expansion in Brussels and POLITICO's creative agency POLITICO Focus during his six years with the company. Prior to POLITICO, Bennett was co-founder and CMO of Hinge, the mobile dating company recently acquired by Match Group. Bennett began his career in digital and social brand marketing working with major brands across tech, energy, and health care at leading marketing and communications agencies including Edelman and GMMB. Bennett is originally from Portland, Maine, and received his bachelor's degree from Colgate University.

Enterprise

Why large enterprises struggle to find suitable platforms for MLops

As companies expand their use of AI beyond running just a few machine learning models, and as larger enterprises go from deploying hundreds of models to thousands and even millions of models, ML practitioners say that they have yet to find what they need from prepackaged MLops systems.

As companies expand their use of AI beyond running just a few machine learning models, ML practitioners say that they have yet to find what they need from prepackaged MLops systems.

Photo: artpartner-images via Getty Images

On any given day, Lily AI runs hundreds of machine learning models using computer vision and natural language processing that are customized for its retail and ecommerce clients to make website product recommendations, forecast demand, and plan merchandising. But this spring when the company was in the market for a machine learning operations platform to manage its expanding model roster, it wasn’t easy to find a suitable off-the-shelf system that could handle such a large number of models in deployment while also meeting other criteria.

Some MLops platforms are not well-suited for maintaining even more than 10 machine learning models when it comes to keeping track of data, navigating their user interfaces, or reporting capabilities, Matthew Nokleby, machine learning manager for Lily AI’s product intelligence team, told Protocol earlier this year. “The duct tape starts to show,” he said.

Keep Reading Show less
Kate Kaye

Kate Kaye is an award-winning multimedia reporter digging deep and telling print, digital and audio stories. She covers AI and data for Protocol. Her reporting on AI and tech ethics issues has been published in OneZero, Fast Company, MIT Technology Review, CityLab, Ad Age and Digiday and heard on NPR. Kate is the creator of RedTailMedia.org and is the author of "Campaign '08: A Turning Point for Digital Media," a book about how the 2008 presidential campaigns used digital media and data.

Latest Stories
Bulletins