14 The Pulse Civic Engagement Platform
Civic Tech Chat | 2018-07-22 | 52:19
We dive into a discussion with [Drake Hougo](https://www.linkedin.com/in/dhougo/) and [Sean O'Bannon](https://www.linkedin.com/in/sean-obannon/) Co-Founders of Pulse, a platform meant to help improve civic engagement between consitutents and legislators.<br><br>##### Music Credit: [Tumbleweeds by Monkey Warhol](http://freemusicarchive.org/music/Monkey_Warhol/Lonely_Hearts_Challenge/Monkey_Warhol_-_Tumbleweeds)
Top Keywords
- bill 0.007
- site 0.006
- government 0.005
- pulse 0.005
- tech 0.005
- data 0.005
- civic 0.004
- technology 0.004
- congressional 0.004
- civic tech 0.004
- political 0.004
- summarization 0.004
Transcript
Speaker 0
0:00 – 0:57
Hello. I'm Ryan Cook, and this is Civic Tech Chat, a podcast about the civic technology movement. We seek to harness the power technology has to improve the delivery of public services to people everywhere. Welcome to another episode of Civic Tech Chat. This week, we're joined by Dre Hugo and Sean O'Bannon, students at Stanford and founders of PulseVoter, a platform that's meant to make civic engagement with legislative representatives easier for their constituents. In this interview, we're going to talk a bit about the platform, its design process, and what it might mean for the electoral and constituent feedback process. So let's go ahead and hop right into the discussion and learn more. Thank you to both of you for coming on the program today and and taking the time out of your day to do so. Could you start off by introducing yourselves and telling us a bit about what you do?
Speaker 1
0:57 – 1:20
Yeah. Sure thing. Well, it's a pleasure to hop on. Thank you so much, Ryan. We're excited to be on civic tech tech chat here. So my name is Dre Hugo. I'm one of the cofounders of Pulse voter. I'm a rising junior at Stanford University studying economics and political science. And we've started Pulse as a way to connect citizens to their elected representatives, and I'm joined by one of my other cofounders, the guy who's has built the whole Pulse platform.
Speaker 2
1:20 – 1:39
Hey, Ryan. Yeah. Thanks again for having us on. My name is Sean. I run the tech side of things at Pulse. I'm a rising junior at Stanford as well. I'm, majoring in CS, focusing in AI, and, really excited to apply a lot of the things that we're, like, learning about in the classroom setting at school and so, like, real world applications that hopefully can help a lot of people.
Speaker 0
1:39 – 1:57
Now one of the things that civic tech chat tends to start out of the gate with with these conversations is all about why, and starting with the the why, one kinda goes to start an endeavor. So could you both talk a little bit about what your personal why is as far as what gets you out of bed in the morning to do what you do?
Speaker 1
1:58 – 3:16
Sure thing. So pulse originated from really just a general frustration with the communication system that we have placed in American government right now. Looking at the current tools we have, whether it's phone calls, emails, letter writing, it feels outdated. And, honestly, people just don't have a voice. And I think if the twenty sixteen election proved anything, it's that people are looking for new ways to engage, and technology is certainly at the forefront of that. My dad back home in, Southern California is a junior high science teacher. And my senior year of high school, we were talking a lot about the political system and how we thought it could be made better. And he decided to run for US senate in California, actually, on a tech platform called Democracy two point o, where he pledged to vote majority consensus of California voters. And it was really my dad and I on this big impossible endeavor, running this out of our garage. We spent less than $3,000 and managed to garner around 65,000 votes in the California primary election. And while we had honestly no idea really what we were doing or how to properly run a campaign, we think the idea of bringing technology to political communication is really, really powerful. And I entered Stanford, met my founding team, including Sean, and we really began on this endeavor. How can we use technology to better connect government? So this has been long in the making for me, and it's really excited to, see the progress that we've made over the past year or so.
Speaker 2
3:16 – 4:15
Yeah. Definitely. And on my end, I originally met Drake through, a running club on Stanford, who we're both, like, very involved with. And through that club, we became really good friends over the year. He started getting really involved in politics, specifically this kind of feedback loop with representative. And he approached me about working on something, on the tech side of things, maybe last year or so. And, since my specialty has always been in kinda developing the iOS apps, web apps, I thought this was a great project. I too had noticed the kind of lack of technology in the political space, and I was, kind of frustrated by that in regards to interacting with my own representatives. And I thought that this would be a great way to kind of, apply, these more theoretical concepts that we learned in school and, the the rising tide of technology into this really, this laggard space. And that's, kinda what has driven me to, invest in Pulse.
Speaker 0
4:16 – 4:57
Now what what I'm hearing there is that it sounds like, you two have some pretty complimentary sets of skills for what you're trying to do here. Drake, I I get the sense that you have a lot of, exposure to the kind of political science, side of things both with your experience with that race and with your studies. And, Shauna, I'm I'm getting the sense that you have a a a pretty, interesting amount of experience on the computer science side, but both with your studies and other works. I I believe you even, managed to publish a few apps to the app store yourself. And I what I would ask is so far, how has those two kind of sets of skills managed to to work to work, for each of you as you've worked together on building this platform?
Speaker 2
4:59 – 5:46
Yeah. You're completely right in that. Well, we do tend to complement each other. Our working styles tend to be very similar. We we've invested, a significant portion of our hours into this project, over the summer, you know, hundred plus hour weeks and whatnot. And we spend, like, you know, almost twenty four hours a day with each other. And in terms of working together, I think that our skill sets play each other well. We kind of decide, higher level goals, together as a as a founding team, and then we kind of break off into more specific tasks. I I can handle the tech side of things, and also, like, our development team and our engineering team, whereas I think Drake does a really good job interacting kinda with the, the outside world, especially the political world. Can you talk a little bit about that?
Speaker 1
5:46 – 6:28
Yeah. Definitely. That's one of the things that especially in this space since we're simultaneously building a consumer product as well as a product and dashboard for congressional offices and campaigns to use. Sean and I and the rest of our team, our two other cofounders, Andrew Cork and Jackson Eilers, they both study computer science as well. I think we complement each other really well. Fortunately, Stanford's political science department has been really awesome in terms of getting us high level introductions to legislators in California and throughout The United States. So it's certainly, my mission to get as much feedback as possible from the congressional level and funnel that into Sean and Jackson and Quirk as we develop our products and build something that people love. So the balance is really key.
Speaker 0
6:29 – 6:53
Now a question that, I tend to ask, on these episodes that at least lately, that that's, been a little repetitive, but I think it's one that's important is something that comes across as a very simple question, but everyone seems to have their own twist of an answer to it. And that's the question, what does civic tech mean to you? So what what would each of you have to say about, if you were to answer that?
Speaker 1
6:53 – 7:58
Yeah. So what civic tech means to me is the future. Civic tech to me is the future in this sense that got I I always like to say the government is the final frontier. So we're you know, the government's living in the twenty first century using twentieth century technology with, you know, eighteenth century, principles and guidelines. Government has consistently been a laggard in the willingness to adopt technology, so we've really sought after congressional partners who are willing to adopt technology as a way of the future. You'll get countries like Estonia, who you would just never suspect to be, like, way more progressive than The United States, and I think they were one of the first countries to have blockchain voting. So, I mean, just little examples like that where you put public utilities online, where you make it a lot easier for people to engage with government. Civic tech is about the future and what the future of government, not only in The United States, but what the world potentially looks like. We you know, most people in The United States have Internet, have some sort of, device, whether it's through a public library or their own, and it's time the government caught up and, meets the people where they actually are.
Speaker 2
7:59 – 9:27
Yeah. Definitely. I completely agree with that. In terms of, like, what the the specifics of this technology looks like, we see, like, a huge, onslaught of of data coming our way in the tax base, and its applications in government, are sometimes thought of more nefariously than I think they should be. I mean, the the force of data can be a force for good, a force of, reaching out to your representative, letting them know what they they feel. And, working with different campaigns, different, representatives, to to bring data into the fold in a positive way to effecting a positive change, has been, really exciting when these individuals that we work with are innovators and willing to adopt something new for the possible positive change. And, at the same time, we still see a lot of people in the space who cling to the more traditional methods. And, from my perspective, it's a little bit disappointing to, to not, kind of endorse this technology for fear of, the the negative consequences. Because I think if we can do, something positive in this space, the that as a society, we're moving there anyways, and if we can control it and, bring some kind of if we can do it in a in a positive way, I see a lot of, net benefits for our government.
Speaker 0
9:28 – 9:48
So with the end of that question, I I'd like to go ahead and have us start to shift gears, into talking a bit about, the Pulse Vitter products that we've been hinting at here in our conversation. For someone that has zero exposure to it, could we talk a bit about, like, what the overview of what Pulse is all about?
Speaker 1
9:48 – 11:42
Definitely. So the one liner we like to use is Pulse is a web platform connecting citizens to their government. So from a really high level, we see ourselves as a digital bridge between the average American citizen and legislators. So the current workflow as it is now, we're currently parked at pulsevoter.com. We're releasing an iOS app in a few weeks this summer. The way it works is as the bills in congress, think border separation, gun control, health care, the economy, whatever it may be, we produce quick hitting nonpartisan summaries on a specific piece of legislation. Citizens going and visiting the site can look this information over, learn about the issue. They can click support or oppose, add a comment justifying why they feel about that way, and all that information is sent directly back to their elected representative where we're working with a handful of both congressional campaigns and incumbents throughout the state California for the for them to really look over and analyze this data and information. And once campaigns and legislators have this info, we make it really easy, and we're looking at a couple different iterations of messaging so campaigns can immediately respond to constituent, explain that they've been heard, and really complete the feedback loop that is so lacking right now in government of giving people a voice, allowing them to participate on issues time and time again. Because if we look at political communication right now, I mean, we have one of the worst voting, participation rates in western democracies. It only happens once every two years. People are just completely unplugged from the process even when voting is going on. So our goal is to reduce that feedback loop from a voting cycle every two years to every few days people are engaging with the issues that people are facing in the federal government. And we fully intend to, to basically inter integrate the government as much as possible. We want people to have a one stop shop to get all their political news, interact with their political officials, everything from city council to congress. And Sean can touch a little bit more on what our product looks like from a tech standpoint.
Speaker 2
11:43 – 12:18
Yeah. So I don't know how, like, familiar, our our listeners might be with some of the more behind the scenes stuff of website, but, we built it with the help of AWS, running, like, one of the most modern stacks out there. And, security, for one, has been one of the most important things for us from a tech standpoint. I mentioned data earlier, and one of the most, complimentary things to data is the securing of that data. So, in terms of technology, that's been one of our core focuses, and, you know, we're using a lot of state of the art methods to achieve that.
Speaker 0
12:19 – 12:34
Now poking around the app myself, one of the things that stuck out to me was this, kind of headline, that was on there in the about page that says our political system is broken. What does that statement mean to each of you?
Speaker 1
12:35 – 14:01
Definitely. Yeah. So I was the one who originally penned that, and what it stems from is a lack of voice from the American people. Our political system is broken because both political parties, in my opinion, have really stopped representing the best interest of the American people. And you can chalk that up to a whole bunch of things, the increase in polarization, the increase of money flooding into the system post Citizens United, the change of the presidency. To me, none of those really matter as much of the fact that government has not caught up with the rapid scaling of technology, and it really needs to. If, you know, you look at the twenty sixteen election and the way campaigning was done and everything, Digital media, online communication is the way, certainly, The United States is setting, and it's really, really important for The United States to pick up on that. Think about how many problems would be solved, how much better Americans would feel about their state government or state and federal government if they just feel like they had a voice. It doesn't matter based on party. We're talking to Republicans, Democrats, Independents, Green, you name it. People feel disenfranchised and just left out of government. And, you know, the the congress approval ratings show that. You know, congress dabbles in between, you know, an eight to 18% approval rating. It has almost about a 95%, reelection rate. Something that's seriously misunderstood between, you know, the high ups in government and the average citizen. So we we just start trying to use technology to connect the two.
Speaker 0
14:01 – 14:24
So I I imagine there's two sides, to this, but I'm curious to hear a bit about what the process is like for a member of congress that ends up being interested in signing on to the platform. I imagine there's some amount of onboarding that is involved in kind of getting them through from a process perspective as well as I imagine there's some, technology element to that as well. Could we kinda delve into what that's like?
Speaker 1
14:24 – 16:33
Yeah. Absolutely. So for anyone listening to this blog, I think the the the trope is that government is inherently behind on technology, doesn't quite understand. I think Mark Zuckerberg's, congressional testimony revealed that when a senator asked him how they made money, and, Zuckerberg kinda smiled and said we run ads. There's a real lack of understanding about technology and government. So we've really chased after who are the young tech savvy legislators in the house of representatives. And fortunately, through Stanford University's political science department, we came across, he's age 37 and just a real innovator on the tech in congress front. He was one of the first people to use Snapchat as a way to communicate with constituents, which is a phenomenal and really clever way to give people a look into the daily life in congress. Our product right now, not every legislator is into, whether they don't understand the value and importance of data or they're a long time incumbent and they don't think they need more tools to connect with constituents. So we're starting out with a pretty small pool of individuals who are young, tech savvy, pretty progressive in this regard. Onboarding for us is very, very simple. Essentially, what we've built is a, campaign and office dashboard that visualizes all the information we collect from the user site. They can log on there, and what they can immediately begin doing is drilling down into the different demographics of their district. So how do women ages 18 to 29 feel on this immigration bill? That really allows them to get a better sense, a pulse of their district for lack of a better term, on where people stand. So for congressional partners right now, we're launching with, which is a really exciting moment for us in this company. Basically, our whole onboarding process really just involves demoing the dashboard for them, showing the power and functionality, talking to them, getting their feedback, understanding how they wanna communicate better with constituents. And then from that, all we're real we're really looking for from legislators is anytime they take a stand on an issue or promoting it out across their social media, email list, and channels, They simply include a link to our site encouraging voters to go participate. And from there, we really hope to drive, especially among the hyper politicized people, a pool of individuals who wanna engage with government.
Speaker 0
16:34 – 16:50
And I I believe you started to allude to this a little bit, but, I would like, to maybe explore in a little detail the the types of data, that that member of congress gets access to as people answer questions. Could we talk a little bit about what sort of data they're looking at and, how that's structured?
Speaker 1
16:51 – 18:05
Yeah. Sure thing. So our main, utility of the site, we wanna provide them information that's valuable to them. So on the user facing site, all we're collecting is first name, last name, ZIP code, and an email address. This is basically so the ZIP code, we can identify where their elected representatives and email addresses so that they can create an account. From this, what we do is we match the first name, last name, and ZIP code to the public voter role and information, and we aggregate that in the back end for the rep dashboard. So from there, we also include information including, like, home address, gender, date of birth, race and ethnicity, and then it also includes facts like, did they vote in the twenty sixteen general election primary? From there, we're still working on figuring out exactly what data analytics and predictive analysis they want, but we're giving them a high level, overview of how people feel on the issues. And the real value prop to both campaigns and legislators is from what we've seen, there's no dataset that includes how individual people, individual constituents feel on specific political issues. So if you're able to go door to door canvassing or, you know, are voting on a bill and you know exactly how someone in your district feels, that's a game changer in terms of the way people, interact with politics from a legislator's perspective.
Speaker 0
18:07 – 18:34
Oh, that that absolutely would be a game changer. I can recall from my own experience going door to door that oftentimes that's a bit of a a blind endeavor. You might have an idea of, like, basic demographics of what someone's like that lives there maybe if if the information's accurate. But those those conversations pretty much always end up being an exercise of essentially trying to figure out what you've already collected there. That that must be a a rather powerful thing to have, for that member.
Speaker 2
18:35 – 19:30
Yeah. Definitely. Additionally, one of the tools that we're, developing right now is kind of like a a canvassing heat map. Basically, breaking down, visually on a geographic, like, map, at at what houses, like, an in like, let me rephrase this. Breaking down, at each house, who lives there, like, what what they believe in, like, on a per bill basis and a per issue basis as well. And we're also allowing representatives to export, a geographic area around, specific, location. For instance, if they're hosting a coffee site. If they're hosting a coffee site chat, they can export maybe a five mile radius, of all the users around their coffee house and say, hey. You know, I'm meeting at this coffee house. Please join me, and let's talk about these issues. And we think that'll be a really effective way of mobilizing, different communities, on the representative side.
Speaker 0
19:31 – 19:44
If if I'm hearing you right, it it sounds like that that heat map, portion that that that is is coming might even be something that could be a significant part of a campaign's field operation. Am I hearing that correctly?
Speaker 2
19:45 – 20:21
Yeah. Definitely. So one of the things we're also working on right now is integration with PDI to allow, different campaign management systems to, basically import this, like, per bill data into their existing systems. And we think that that'll be a really powerful tool for them, especially on canvassers. You know, for instance, a canvasser could walk up to a house where they know, that an individual supports, you know, a a bill, a certain bill, and ask them, you know, hey. I saw you that you voted, you know, yes on, for instance, keeping the families together. Did you know that, yeah, reports this, as as an example.
Speaker 0
20:22 – 20:34
And out of curiosity, as a follow-up to that, I I I believe you mentioned that you're trying to integrate with a couple of other different tools. What has your experience been like trying to integrate with those existing, players in the space?
Speaker 1
20:35 – 21:31
Yeah. So that's something that we that that honestly is pretty new for us right now in the sense that we are we are customer and use user obsessed. We're we're out in the streets talking to voters. We're trying to talk to as many campaigns as possible. And, really, we've gotten a lot of great feedback from campaigns, and that's something that we quickly realized is that our tool is a great supplement to some of the other resources and whatnot they're using. So PDI integration was something we quickly realized that we needed to have because they wanted to use our data in a workable format for canvassers. Other tools that, they were looking at include, like, NGP van, which is, you know, fundraising and direct email marketing sites. So we certainly wanna be the channel to you know, for users to constitute or to communicate with elected representatives. But at the same time, we wanna provide the data that's gonna be in the most valuable form for campaigns and official, offices as well.
Speaker 0
21:31 – 21:47
It's become clear that essentially, like, this this tool is made, not just necessarily for existing, office holders, but also for those who might seek office as well. Is is there any difference in the experience, depending on which side of that fence you're on as far as, like, if you're interacting with false?
Speaker 1
21:48 – 22:51
Yeah. Without a doubt. And this is something that I think a lot of civic tech founders and whatnot can really relate to the struggle of, and this is something that we're working through right now too. Because of the way the laws are set up and for good reason, there's a very strict wall in between the official, federal office. So as of right now, we're working primarily on the campaign side with the hopes that we can prove value there first and then integrate well with the offices. This involves undergoing sort of the government procurement process and a whole range of other bureaucratic red tape that we had to cut through. But from the high level vision of what we see being important to American democracy, it's important that we're in actual congressional offices. You know, we we know we'll have made an impact when a legislator is able to walk on to the congress floor and read a comment from a constituent in their district or, you know, read off one of the polls we ran and use this as a source of information in their decision making process. So we're we primarily live on the campaign side with the full intention of being able to prove value here before moving into the actual offices.
Speaker 0
22:52 – 23:27
It it's interesting you bring up, the government procurement process and kind of those processes in general. A question that kinda came to mind to me as as you were describing that, kind of revolves around compliance, and it's something that I I maybe you might know the answer to that that I don't. As you work with individual, office holders offices, do they have different, I guess, standards for compliance with how you treat data, how you handle it versus other offices, or is that something that's kind of standard for if you're going to manage data for congressional office, it has to meet these particular standards?
Speaker 1
23:28 – 24:09
So certainly, in the house of representatives, absolutely. On the campaign trail, it's really kind of the wild west, so we don't have to worry about any sort of compliance issues on that front. In the federal office, obviously, even though they're a partisan you know, they're running for a position with a partisan stance, They're supposed to represent their entire constituency. So there's a lot more rules and regulations around there. That's why we're starting specifically with campaigns simply because we need to find product market fit before we jump through all these loopholes or not loopholes, before we jump all these legislative hurdles in order to get this into offices. So kind of vague, but, I I hope that adds some clarity to where we're at.
Speaker 0
24:10 – 24:17
Yeah. That makes sense. It sounds like it's a a bit of an adventure, yet to be had, once you're able to to to get off the ground with your existing efforts.
Speaker 1
24:18 – 25:04
Am I hearing you right there? Yeah. Exactly. So sort of where we're at right now is we're running, starting next Wednesday, about a five week pilot program and really trying to prove technology can be used as a viable form to connect voters and elected representatives. From there, we plan on just opening the floodgates and expanding our, legislative Rolodex, bringing more incumbents onto the site and platform, and really just scaling a product that people like and can engage with government with. Going back to your question earlier about, you know, what is civic tech? Civic tech is a way for people to be involved with a low barrier to entry because not everyone can go out and protest. Not everyone can run for public office, but most people already have apps on their phone. They have access to a computer. So let's meet them where they're at with the product that's, simple and easy to use.
Speaker 0
25:05 – 25:31
Looking through the app, it it seems clear to me that there's a lot of intention and thought that's been put into the the data sources. You're pulling to give folks information about the individual bills they're voting on. I believe you're using, congress.gov, the congressional research service, the congressional budget office, or or the CBO, and gov track. Could we discuss a bit about, like, what went into the decision for those specific resources?
Speaker 2
25:32 – 27:40
Yeah. Definitely. So the resources we chose, those about five that you just mentioned, were chosen because they were, unbiased and nonpartisan. None of those sources, at least explicitly contain a bias or upfront about having bias. Whereas more popular news articles, like CNN, Fox, even Breitbart, have bias. So we wanted to provide our our users with, an effective means of education before they vote on a bill, just because that's something that's important, not only to, the representative because they need, insightful feedback and meaningful feedback, which they can only get from an educated voter base. But also, from an ethical standpoint, we think it's really important that our voters, understand what they're voting on. Otherwise, our democracy becomes a little bit less intelligent. We we have a higher probability of or have a lower probability of, you know, making the correct decision on on certain bills. Yeah. So I guess it also in terms of, like, going forward, we're doing a lot of this, like, data collection right now manually. Drake, way back in the day was doing this by hand, you know, staying up till, like, 2AM in his dorm room, writing these summaries, and then we would copy and paste them into our database. We're a little bit more advanced right now. We have a team of writers, looking at some of these, these sources, commerce.com, GovTrack, Libko, and creating, summaries from there. Ultimately, what we hope to do is develop, a summarization algorithm. If not, you know, for complete summarization of bill and, bill texts, then at least a tool to help our team, more effectively summarize what they're doing. So this could look like, you know, extraction. Think of an automatic highlighter, basically. An automatic highlighter that goes through a, text of a bill and picks out the most important sum summaries sorry. Automatic highlighter that goes to the text of a bill or summaries about a bill and picks out the most important census to include, to most effectively educate our voting base. We think there's, some really cool tech just waiting to be applied to this and, some really cool applications.
Speaker 0
27:41 – 27:59
It it sounds like that that process of trying to figure out what sentences have that level of importance, is a bit of a unique challenge, perhaps an interesting one to attempt to solve. Do you have any insight so far as, like, what sort of challenges might come up in, like, trying to to make that happen?
Speaker 2
28:00 – 31:06
Yeah. Definitely. So, text summarization is notoriously difficult and definitely still an unsolved problem in the space of, like, natural language processing. So we're reaching out to a lot of our professors. Now. We're we're taking, undergrad and grad level classes and NLP at Stanford, and we're making, really effective use of the contacts we made in those classes. There's effectively two ways of, summarization, like full text summarization. There's just abstractive, which goes through a text and picks part, like, base basically converts the entire text into a, quantifiable, format. Basically, a string of numbers indicating, like, meaning. And then from those numbers, tries to generate text. And this is abstract of summarization. But the way we do this, we're limited by, some of the architectures that we use, and we can only really generate, you know, a sentence or maybe two from a document of a couple 100 words. And so for a bill that could be upwards of 800 pages long, think like ACA, this isn't feasible. So, in terms of what we're doing right now, there's, like, two ways, to go about it, the the extracted way. There's basically pre neural, which is, like, pre neural networks, pre, what we would call, like, AI and the, third neural way, which is more along the lines of AI. And in terms of pre neural, we there's one popular algorithm called TF IDF, text frequency inverse document frequency, which looks at the, lists of, senses and finds the, the senses with the most unique words and basically pulls those words out and says, like, the senses that have most unique words are probably the ones that are the most important. And then the neural way would be to basically train, a neural net to classify sentences into levels of importance or not importance. And, one of the biggest challenges for us in the the neural field and basically the the abstract field too is that the data for this is quite limited and, doesn't really exist in the capacity that we needed to to do some of the more, advanced, like, techniques. So, in terms of, like, where I think the, the technology is gonna go and what, we're initially trying to collect in these, initial few months is, a big enough data set such that we can really begin applying these more, advanced techniques, these more advanced models, and learning a lot about, like, even what the the semantic meaning of a bill is and, you know, what are some features that of a bill that, like, make it so unique. And, you know, if there's a limited domain of a bill, you know, can we use that to our advantage to make more effective summaries? So, we're really trying to learn a lot about bills, summaries, and, really trying to reach out to all our contacts and, really apply some of these state of the art features to bill summarization.
Speaker 0
31:07 – 31:38
That sounds really fascinating. I I imagine, trying to do this for legal language itself is a little different than, trying to do so for, say, a a a work of fiction that was written. In that my understanding is that for legal language, there are some statements that have specific meanings, in a way almost like a programming language does that are are used for for a specific structuring within a document. Does does that sort of thing kinda play a role in in what you're attempting to do with those techniques?
Speaker 2
31:38 – 32:28
Yeah. Definitely. And, the, like, legal language side of things makes it really problematic to summarize just, the text of a bill. In terms of presenting, information in the cleanest way possible to our users, that's why we make use of those, sources like GovTrack, ProPublica that we mentioned earlier because they tend to take some of the legalese out of the document or the bill. One of my favorite examples of, you know, what what, the summary can cover that is not immediately apparent in, perhaps the text of a bill itself is the financial cost of bill and the financial repercussions of passing it. And, doing like, bills often, don't contain that language. And what is actually required is a thorough analysis of a bill by a government agency.
Speaker 1
32:28 – 32:40
I think Drake knows which Yeah. The Congressional Research Service does all the they do all the research by hand right now. We've actually talked to them. They're really interested in automating at least some of this process.
Speaker 2
32:41 – 33:24
Yeah. So in terms of summarization, it's not, like, feasible within the next, I'd say, couple years or so to do a complete summarization of a bill, but we think it's definitely feasible to, you know, with this automatic highlighter pick out sentences that are important and assist, any kind of summarizers, whether whether it's the Congressional Research Service or our own team of writers in, creating effective summaries for, the general public. Because that's another thing that we're seeing is that, the the voter mass, we just need to be better educated so that we can make more effective decisions on the bills. Especially in a post back world, the the more effective, accurate, summaries we can provide to people that they truly believe, we think the better our democracy can evolve.
Speaker 0
33:25 – 34:35
Yeah. I I think you you make a really great point there, in particular that the statement about we we need to know more about these things. I know that, in in some reading I've done in the past, it's become clear that e at at even in, like, state legislatures, there's, like, very different levels of professionalization that can occur. And I imagine there's probably some impact on that in congress as well, just even simply based upon the person that's in office. So I I find it encouraging, that that that you two are in in your the rest of your team are are working to improve that situation, so that folks can be more knowledgeable about these things going forward. Also related to the the product building side of things, I also kind of in looking around at it, it seems that there's also significant amount of intent in just the way it's laid out and designed. For example, I'm a bit of a fan of the way that you're able to kind of very quickly get that bulleted deal, that that you've, that you were kinda talking about how that's generated, as well as, then immediately be able to get to the full text if you're interested. Can we talk a little bit about the design process that you went through and kind of figuring out, like, what how should Pulse be presented, to the citizen coming to it?
Speaker 1
34:36 – 36:06
Totally. So this really stems from, once again, talking to users. So we've gone through a bunch of different iterations of the site. We have a whole graveyard of different sites, some that had graphics, one that was called votehero.org. And what it's really stemmed from is what's the best way to provide information so that, say, my younger sister, who's a sophomore in high school, is able to understand a really complicated piece of legislation. And, fortunately, we've had the privilege of working, with a phenomenal designer, named Vincent who's helped us really just sort of shape our our whole product and UI and layout. So, I mean, our main goal, if you if you land on the site, we have cards in a newsfeed format. I mean, if you look at any sort of modern social media app, Twitter, LinkedIn, Facebook, they all have the newsfeed style. Off these cards, we wanna give a high level summary. What does this bill do right off the bat? From there, you click read more, and we wanted to give you just a quick few bullet points on what the piece of legislation is. And, you know, we've actually shown this to a few congressional offices, and this has been one of the pushbacks. They're they respond, you guys aren't providing enough in-depth detail. We don't really wanna know how people feel, on this issue if they're not, like, fully informed. My response to that is, how many legislators are even reading these full pieces of legislation? For the most part, people are voting with, you know, with their hearts and their emotions as well as their mind. But if we can give people a high level overview of what bills include and what they're doing, we see that as winning because it's currently better than the status quo, which is most people don't know anything about what's going on in government.
Speaker 2
36:07 – 38:11
Yeah. So I think in terms of, like, how people vote on bills, people have, I think, generally, their their values, fairly, like, constant, and that they they don't change that much. But I think, one thing that we can do a lot better of, as a society is, like, phrasing bills in a way such that we can apply our values to them, and that's really what we try to do with our site. So, I think most people know how they feel about, like, for instance, you know, controversial issues like the environmental gun control. And we just wanna present a bill that might touch on these issues in a way that they can clearly, apply their values to it. Back on terms of, like, the design front, we, we we put a lot of effort into how, people first interact with the site and what they what what it brings out in them on, like, the their first, visit and their first experience. So we try for, a lot of brand consistency across the site. We have the, the cross platform app coming out soon, and so we're trying to use a lot of, like, similar similar interactions between the website and the app. The colors we found, through user testing, to be the most successful are the ones that are on the site right now. So and we definitely encourage the listeners to go, give it a visit at www.pulseloader.com. We you know, my experience comes from, iOS apps, full stack development. And having a really clean product is something that I always strive for. And, you know, I use this product myself too, back when I was, you know, just frustrated at a couple things happening in my home district. And I I looked at the the site and, you know, tried to, you know, make like, see how a frustrated user would interact with the site. And so by testing not only other users but also our own interactions with the site, we're consistently iterating and validating and learning from the site itself using this agile framework development that is so popular with startups these days. Really trying to to make a difference, on a micro level and thus on a macro level with our own property.
Speaker 0
38:12 – 38:28
So if there's a, say a member of congress out there that is a secret civic tech chat fan, and they're hearing all this, realizing that Pulse is available now, and they're like, man, I I want to hop on this platform. How would they need to go about getting that ball rolling?
Speaker 1
38:29 – 39:23
Without a doubt. I mean, if they're a congressperson or even civic tech enthusiast, please reach out to us at at the email address contact@pulsevoter.com. What we're doing right now, and especially, this is one of the advantages that we have being a a, you know, pretty small student team and just being a sort of startup in general. We move really fast. And tapping on what Sean was saying about our sort of agile development framework, we can pump out different feature sets, big changes to the site almost every single week. So if anyone is looking to engage with us, please reach out to us, and we'd love to figure out a way to work together. Our whole goal at this point is to prove that this product can work in at least one geographic area. Because if we can get one of the 435 house of representative members on this platform, we think, well, maybe a slow and tough process, we can slowly pick people off one by one until this is the gold standard of political communication.
Speaker 0
39:24 – 39:34
And in a similar vein, if there are folks out there that would like to lend a hand in in in the mission you're doing, is there a way for them to do so? Are there things that you could use help with?
Speaker 1
39:35 – 40:11
Yeah. Definitely. I mean, we need people who are engaged in government, who are frustrated, who are happy, who just wanna have a voice. Come to the site, engage, vote, and share it out with friends and family who are also engaged or looking to better learn about the political system. If you come from a technical background, we we definitely have internships and positions. We're doing a lot of really exciting data analytics work around voter data and how it can be used by campaigns and legislative offices. So if you're looking to work on a company that's trying to redefine political communication over the next fifty plus years, Pulse is the is the company for you.
Speaker 0
40:12 – 40:26
Also, I I I recall now that in the conversation we had, before we started recording, you mentioned that there's a, a new feature, coming out on Monday that is something to do with user to user interaction. Could you are you able to talk a bit about that here on the program?
Speaker 1
40:27 – 41:55
Yeah. Absolutely. So thing that we're really excited for. So as of right now, Pulse is set up so that there's a communication channel between user and elected representative. But what we realized is outside of geographic areas where we have a congressperson or campaign there actively promoting it, there's really not much of a reason for someone else to come on to the site and engage. So what we were really looking at, especially in these really divisive times, how can we get two people to interact through sort of a a messaging feature on the platform? So on the on the site, we're unveiling Monday, a messaging system that essentially you go on the platform, you vote on an issue you care about, you take a stand. I support stronger gun control rights. And what we do is we randomly match you up with someone who doesn't, someone who opposes that viewpoint and allow you guys to talk one on one. What we see a lot of times on Facebook, what Reddit, these big forums, it turns really nasty. People are trolling. They're throwing insults. They're throwing big facts and different information out there. When you sit down with someone and you have a one on one conversation, typically, it's a lot more productive because it's less confrontational, and just I mean, more productive generally. So sort of our four rules when people enter the site, you're gonna agree on some things, you're gonna disagree on other things, keep an open mind, and we're gonna walk away having learned something. Those are the principles that we think are not only important to America right now and in the future, but are gonna really allow our product to be successful. Yeah. And in terms of, like,
Speaker 2
41:55 – 43:15
you know, how some people might be very pessimistic about this. They think it it may not work. People may not even, like, wanna communicate with another. We don't see this as, you know, a tool to communicate, with the the other side, but more of, like, a solution. We wanna walk you through each step of getting to know the other person, getting connect to connect with them on an individual level, which we think actually, leads to greater successes and connecting over something that is typically very controversial, like, gun rights or gun control. And, in doing so, we think that we can allow people to find common ground on certain things. We ran, one beta test over video, earlier this week, and we discovered that, two individuals, one a democrat from this area and one a republican from, was it Minnesota? Mhmm. Yeah. Connected over, gun rights and gun control. But they came in with very different experiences, and they found that, one one thing they learned was that they both supported that, guns should be less lethal in America. And they they created a couple different, you know, solutions to do that whether by limiting, like, gun capacity or certain types of guns. But, that that was their solution. We think that, you know, individuals could connect and, like, see see where each other are coming from over, different issues as well and propose various other solution.
Speaker 0
43:16 – 43:41
I I I like the sound of that. It it it seems like by, breaking it down to a one to one interaction that you essentially mitigate that dog pile effect that you get from, a site like Reddit or something where it's a giant forum. I am curious though, like, what what is the, the length of interaction that that's intended, as far as, like, how long are two random folks connected? What what does that look like?
Speaker 1
43:41 – 44:11
Definitely. So, I mean, we've modeled our UI off of similar messaging apps like Tinder and whatnot. Quite honestly, Ryan, we're we're testing this out. In our video chat demos, the the two participants talked for forty five minutes straight and were upset when we cut them off with five minutes to go. We're not sure if this same level of energy and enthusiasm is gonna hold over text messaging. But, I mean, we're hoping people are in these conversations for, you know, at least twenty minutes or so, and and that's really the ultimate test. Like, can you hop on? Can you talk to someone? Build empathy.
Speaker 2
44:11 – 44:38
So it's a grand experiment we're looking at trying. Yeah. Ultimately, we really want people to stay on until they find the place where they find common ground. And that's how our UI UX is set up right now where, to lead the conversation, you actually need to have found common ground first. Otherwise, it'll just keep going until you can find something that you actually agree on. Oh, that that's really interesting. So the the interaction is is more or less controlled by finding that common ground and the the two parties deciding,
Speaker 0
44:39 – 44:46
okay. Like, we're finished with this right now. It's not so much like there's an opposed time limit sort of situation. Am am I hearing you correctly?
Speaker 2
44:46 – 45:13
Yeah. Yeah. So, when we were talking to our users, that's generally what we've, figured out. Basically, it something else in the user's life could come up, whether it's a phone call they need to, like, leave the chat, but they could always come back to it. And, you know, when they come back to it, we want them to stay there up until the point where they've, they found something. We think that's the point at which the conversation becomes really meaningful, and our users have validated that as well.
Speaker 0
45:13 – 45:41
And, as far as the the matching process goes, something that I would be curious to hear about if if you're able to share is that you mentioned that there's there's, like, some level of randomization, bay but also it's it's in some way based upon their position on an issue, like two people that disagree. What what does that process look like for trying to match up two folks on the chat platform together? Yeah. Definitely. It's a really good question, and it's definitely nontrivial. There's a number of ways we could go about this. Right now,
Speaker 2
45:42 – 46:50
we're congregating everybody who visits the site into basically two different buckets where they can talk about either gun control or immigration. And we think that this will allow people to find, you know, enough other matches around the country that there will be a lot of activity on the platform over, both of those two things. In terms of matching, right now it's random. But one thing we did notice is that that the more people can connect on an individual level, the better they can connect over a controversial issue. So, we think that, connecting people over specific, interests or, demographic lines could lead to positive, results down the line, but that's something we're still investigating and something we're constantly looking for, feedback from users on. So, there's definitely more complex matching algorithm to come, but for now, you can compare it with someone who lives next door to you or someone across the country. And we think that, the level of ramness could also be a good thing. And in the the beta testing,
Speaker 0
46:50 – 46:56
you folks were doing, was there an interaction, AnchorDot that perhaps surprised you the most?
Speaker 1
46:57 – 47:57
Yeah. I mean, it really boils down to the fact that when we've seen two users just chat for five to minute five to ten minutes beforehand and find common ground over music or an interest in a sports team, it is amazing how so something so small and apparent seemingly insignificant can make such a difference in terms of leading on that conversation. I think about two years ago, Heineken did a commercial where they paired two random people up who fundamentally disagreed, a a feminist and, someone who completely agrees or who disagrees with the movement, climate change denier, climate change advocate. And, basically, he sat them down and they drank a beer, really touching a moving commercial. And that's what it builds. Like, we want people to know there's a human on the other end. They have a family. They have a background. They have lived experience experiences that maybe shared or not that contribute to their values and ideas, and that's been probably the most exciting thing to come out of our beta. So be sure to look for it on Monday. It's gonna be an exciting feature, and we're hoping to see a a lot of user growth.
Speaker 0
47:58 – 48:12
Definitely. I I know, hearing you describe that, that that makes me excited to see it as well. So I know I'll personally be checking that out on Monday, and I'll be curious to, learn, what ends up happening there. Perhaps if if you folks are up to it, we could, check back in later
Speaker 1
48:13 – 48:46
and, and talk about what kind of results you saw. Yeah. Absolutely. We'd love to come back on the show. We for everyone who's working on civic tech out there, we we appreciate what you're doing. It's it's at times, it feels like a really small community. You know, David versus Goliath, this old archaic system, trying to bring it into the modern world. But we've met some incredible people through through this process who you know, they're not doing it for money. You're they're not doing it for fame. They're doing it because they, I mean, they love America. They love, you know, this country, and they're pouring time and energy, and their talents into developing better tools for everybody.
Speaker 0
48:47 – 49:01
So as we draw to a close here, one of the things we do on civic tech chat is give the guest the chance to kind of leave us with the parting thought that they'd like to leave us with. So do each of you have some thoughts that you'd like to leave us with as we end the program?
Speaker 2
49:02 – 49:45
Yeah. Definitely. I can go first. I think that, data and political, data and politics has gotten, somewhat of a bad rep lately. I don't think that that necessarily has to be the case. Data can be used for good. Data lets your representative know how you feel on a specific issue. Data, can affect meaningful change at the federal, state, and local level. There's no reason that technology and government has to be so laggard, and I know I'm I'm speaking to the choir or preaching to the choir here, but I think that's definitely a message that, all of us can share and unite behind. Technology and government can be a driving force for good, and we can all get behind that. Yeah. And from my perspective,
Speaker 1
49:46 – 51:22
more on on the political side, we are at a really pivotal moment in American politics. And regardless of your personal, feelings on our current president, our last president, it really doesn't matter. We're on the brink of a huge demographic shift within The United States. Last year, millennials became the largest voting block in The United States population, out passing the baby boomers who unfortunately are are passing getting older with age. And with this change of of young voters, people who've grown up with technology more ingrained in their lives every single day, government is gonna change as well. And while it hasn't been as fast as other industries, whether it be transportation or food or science, Government to me seems like the last frontier and something that the the hard workers now who are putting time in are gonna be making the most difference. We're four months away from the November 2018, general election, which which is gonna have a huge implication and, you know, the the house who retains majority, what legislation is gonna be able to be passed the following two years. And then in the year 2020, when millennials when, you know, my 16 year old younger sister who grew up knowing nothing better than technology steps into the voting booth. It's really gonna be the leaders who adopt technology and meet these new voters in the middle who are gonna be the ones who are prosperous and lead the new way in American politics. So can't thank you enough, Ryan, for having us on the show. We're really excited on what we're working on and just hoping to connect citizens and elected representatives in order to make America a little bit more efficient and a little bit easier for everyone to work with. So thank you to all the other, civic tech listeners for giving us the time.
Speaker 0
51:22 – 51:45
Thank you both for the kind words and for hopping on the program. I, I know myself that I get excited about anything that's all about trying to increase civic engagement, and I know that there's a lot of folks in our audience that that share that view. So I'm sure there will be a lot of folks out there that will enjoy what you've had to say and will have taken something from it. And perhaps we'll, have some sign ups happening there to check out the thing on Monday.
Speaker 1
51:45 – 51:48
Let's hope so. Yeah. Come visit pulsephoto.com.
Speaker 2
51:48 – 51:56
We've put a lot of work into it and are excited to hear your thoughts. Yeah, Ryan. Thank you again for having us on the show. It's been great to talk to you about these things that are so pressing right now.
Speaker 0
51:58 – 52:10
You can follow us on Twitter using the handle at civic tech chat. Visit us on the web at civictech.chat, or subscribe to us for content updates wherever it is you download your podcasts.