I Interviewed the Builders of Boostie. Here’s What They Did.

As the content intern here at Boostie, one of my goals is to be as self-sustaining as possible. Basically, I try to let the team continue to fry the bigger fishes knowing that they are still getting some quality content out every day. For this month I’m doing the complete opposite, badgering some of our guys for a combined 4 hours of interviews.

What I ended with is a pretty damn good look into what we do, how they did it, what’s new, what’s next, and why more people should work with us.

On that last bit, our CEO, Travis gave me the most perfectly blunt statement:

“[You’re] probably wasting a ton of money and you don’t even know it. You’re probably wasting a ton of time and you just don’t know it.”

He continues, “Back in the day, staffing agencies were known for their unique access to talent…at some point everybody went to job boards…nobody had any discernible difference in talent from the next agency over. The difference then became how engaged an agency could be with talent that exists in every other ATS”

Meanwhile, the bill keeps growing. According to SHRM’s benchmarking data, the average cost to fill a non-executive role climbed from $4,129 in 2015 to $5,475 in 2025, up about a third. Chris, our co-founder and Chief Revenue Officer, sees it with nearly every firm he talks to. “Staffing firms end up paying to acquire the same candidates over and over again,” he said. “We’re seeing talent acquisition costs increase year over year, and for many staffing firms, it’s already one of their largest expenses.”

That’s money you’re just chucking out into the void on people you’ve already found. Everything below comes back to one idea: own your candidates instead of renting them. As Chris put it, “If you want to truly own your applicants, you can’t let them live outside your process or experience. You need to direct them toward an asset you own and control.”

I’m laying the rest of this as how a candidate flows through Boostie, from finding your job to being put on a recruiter’s call list. But first, a quick intro to each of our guides on this trek:

• Travis Arnold, CEO and co-founder. Before Boostie, he co-founded HereFish and led marketing at Sendouts, both acquired by Bullhorn.
• Chris Dues, co-founder and Chief Revenue Officer. Leads sales and customer success.
• Kevin Jones, co-founder and CTO. Previously co-founder and CTO of HereFish. Current owner of one hell of a mustache.
• Brian Jones, Senior Software Engineer. Started with a degree in supply chain management and somehow now uses it to catch fake job applicants.*

Additional Note, yes, Brian and Kevin are indeed brothers. Currently referred to as “The Jones Boys”, but R&D is working on something more marketable

1. Boostie Boards: a career site you already own

So Boostie Board (or Boards if you have divisions or other brands) is our own proprietary Career Site, it’s just (legal edit: LIKELY) better than the one you currently use. Launched about a year ago and branded for you, it’s a mobile-friendly job hub on your own website. Jobs flow in from your ATS automatically, with every page set up to show up in Google for Jobs.

While we were talking, Chris was swooning over the AI job rewriting, “I could talk about this for an hour, but we’re talking about a process that should take just a few seconds,” If you were expecting a concise summary, you have not spoken to Chris before. Fortunately, he is a rambler who tends to be right.

The problem is one every job seeker knows. “Job descriptions are too long, and most candidates aren’t reading them,” Chris said. “The most important details are often buried in the middle or at the bottom.” Boostie rewrites each one around what candidates actually look for: the real requirements, the nice-to-haves and the benefits. The target is 350 words or fewer, based on research across thousands of applicants. Each job is SEO/AEO optimized, making it easier to find across Google and AI search tools.

There’s a brand angle, too. “Think of your career site as your storefront,” Chris said. “If it’s outdated, disorganized, or unappealing, you’re turning away potential applicants and clients before they even have a chance to engage with you.”

The ownership piece is simple. Every visitor who lands on your career site instead of a job board’s page is someone you get to keep, and firms typically see more traffic once it’s set up.

2. Convert: an application that works like a checkout

Chris’s take on Boostie’s apply process always tickled me. “Think about the last time you saw a product advertised online and ended up purchasing it. That’s how we think about jobs. The job is the product, and the candidate is the buyer. Our goal is to get the right job in front of the right person and make it incredibly easy for them to apply.”

It all starts with just asking for an email address. “If they abandon the application, we can automatically follow up and encourage them to finish,” Chris said. “It’s just like an abandoned shopping cart when you’re buying something online.”

Travis added: “E-commerce is the most tested and high-converting interaction we have as human beings.” The numbers back him up.

Across all our clients, 91% of candidates who give an email finish their application, and about 80% of people who view a job start applying in the first place. Along with a name, email and resume, the average applicant also answers 3.3 qualifying questions, which gives recruiters better data and helps match the candidate to more relevant roles. For a digital marketing parallel, most marketers are thrilled when 2 to 10% of visitors take any action at all.

Once the basic application is in, Boostie decides what happens next. To move on, an applicant has to clear three checks: no knockout answers on the qualifying questions, no fraud flags, and a fit score above the bar set for that job. Anyone who doesn’t clear them isn’t sent to a screener. “Instead of wasting their time with a screener, we show them other job matches that may be a better fit,” Brian said. Potentially risky candidates are flagged and moved into what Chris calls “a review queue” for further human evaluation.

Applicants who clear the bar get the Advanced Screener right away, while they’re still interested. This is where the questions turn behavioral, some “Tell me about a time you…” kinda stuff. Brian explained: “We use these results to confirm or contradict the score they got from the first phase, helping point out when an applicant’s real-time responses may not line up with their resume.” Candidates can type their answers or talk them through with a voice screener, and firms can have AI help write the questions. Every question, and the score needed to move on, is configurable.

Catching candidates in the moment pays off: one client saw 4.1x more completed screens. Firms that already use a screener they like, including Bullhorn’s Amplify, can plug it in instead.

2 smaller things to note: Candidates who apply to the same job twice now update their existing application instead of creating a duplicate. And firms can set questions once with job rules, which Brian built. “If you know you’re going to need the same question on every job or half of your jobs, you shouldn’t have to go in and do the same thing 200 times,” he said. Rules can be carved up by client, location, job title or category to stop duplicative work and get the right data for the role.

3. Protect: catching applicants that look real

Brian, the senior engineer behind Fraud 2.0, has to spend a lot of time adapting in a resume-based arms race of sorts: people submit fake job applications on purpose, at scale. Some are farms playing a numbers game. Some get hired and hand the job to someone else.

The old way to catch them was checking each email, phone number and IP address against reputation data. That stopped working, because the farms use real ones. “It’s a real person doing the application,” Brian said. “So we can’t just check for a bot and say, ‘No, this is not real.’ It’s the way that they’re applying.”

Travis said it started with clients asking why someone who looked fine wasn’t flagged. This led to “forensics on those applicants,” and eventually to Fraud 2.0, live since August 19. It looks at patterns across applicants instead of one at a time. Brian’s example: “Maybe we got six applications to the same job that have six different emails, but they all have the same phone number. That’s not a real person. That’s somebody using a bunch of different aliases to try to overload a job.”

Before turning it on, the team ran it in shadow mode, scoring applicants without acting on the scores. “It’s a lot of little things that add up,” Brian said. “So we wanted to run it in shadow mode to see what the numbers would tell us, before all of a sudden clients are getting a bunch of people marked as high risk that probably aren’t.”

The results of this are kinda strange, Fraud 2.0 flags fewer applicants than the old system, but with more confidence in each one. And about 1 in 8 of the applicants it does flag would have slipped through before. Put simply, fewer real people caught by mistake, more fake ones caught.

It varies a lot by industry. Fraud is heavy in IT, where jobs pay well and often stay remote. It’s rare in healthcare, where licenses get checked. As Brian said: “It’s hard to fake being an RN in the office.” Recruiters can override any flag they disagree with. Ideally, though, they don’t notice it at all. “Hopefully it’s doing all the work in the background,” Brian added.

4. Review: applicant cards you can read at a glance

Picture a recruiter with a few hundred applicants on one job. Travis designed the new applicant cards for that person, and he has a test for whether a layout is doing its job. “There’s a concept in UX called the blur test,” he said. “If I blur my eyes, would the hierarchy and color of what I’m looking at help me perform whatever action is needed?” On a well-built page, the important things still stand out. His goal for the card was to make the key inputs from each applicant “very easy to digest,” with the full context waiting in the drawer.

At a glance, each card shows how strong a fit someone is, how their AI screener went, their recent role and location, and any trust signals worth a second look. If you need more, each card comes with a drawer. Open it and you get what Travis called “the whole enchilada”: why Boostie thinks they fit, their answers to every question, any knockout answers, the AI screener results and the details behind any flag.

Two choices stood out to me. First, Boostie doesn’t pretend to know everything. When a resume is too odd to read properly, which Travis said happens to under half a percent of applicants, the card says Unknown Fit. “We don’t want to just infer,” he said. “We just want to say it’s unknown. We don’t know. Take a look.”

Second, recruiters stay in charge. “Just like your inbox, some stuff might get flagged incorrectly,” Travis said. “So we allow everybody to override that.”

Cards also show whether another recruiter already owns a candidate. Travis watched recruiters leave Boostie, look a person up in the ATS and come back, just to check. “Those minutes add up.”

The new cards are rolling out to everyone now. And the strong candidates who don’t get the job? “They’re not just going away,” Travis said. Boostie keeps matching them to new roles, which brings us to the next part.

5. Re-engage: the matching engine

If you know Boostie, you’ve probably heard us say something along the lines of this, the best candidate for your open job is probably already in your database. You just can’t find them.

Kevin explained why. Matching used to run on categories. If someone was filed under “software developer,” they got software developer jobs, “but it could have been for technologies they didn’t even have experience in.” Now Boostie compares the person’s actual resume to the actual job. No resume? “We take whatever we can get,” Kevin said, like past jobs they were submitted to or skills listed in the ATS.

Categories turned out to be a bigger problem than anyone expected. For most clients, Boostie’s AI picks the category for every job that comes in from the ATS. A few clients prefer to keep their own. But when Kevin checked those clients’ jobs against the new matching, 25 to 35% were miscategorized. “I didn’t realize that it’d be off by that much,” he said. They likely didn’t either.

Chris explained the challenge: “Let’s say I specialize in placing project managers, but a client sends me a program manager opening. If my ATS doesn’t have a category for program managers, how do you market the job? That’s the problem with relying on categories. You’re forcing jobs and candidates into predefined buckets that don’t always reflect their actual skills or experience.”

Kevin doesn’t blame the recruiters or firms. Filing a specialized healthcare or technical role correctly takes expert knowledge the person setting it up often doesn’t have.

That matching engine now shows up in three places.

Job alerts. Candidates get emails and texts about jobs that actually fit them. “If we include three jobs in an email but only one of them is relevant, do we really want to include the two that don’t make sense?” Kevin said. “We’re going to hold off on showing them junk.”

New jobs. When a job is published, Boostie matches it against your database right away. Chris said the list is usually ready within about a minute.

Talent Pools 2.0. Every job gets a ranked list of candidates from your own database, built from fit, distance and recent activity. Kevin framed the problem: “They have hundreds of thousands of candidates in the database. How do they find the top 10 people who would be good for this job right now?” Distance matters: “Somebody who’s two miles away from the job is going to get a much higher score than somebody who’s 500 miles away.” So does timing: “If they’re not active recently, the most likely conclusion is they just got a new job.”

Recruiters see why each person matches, can filter by distance and status, and can tell who’s already been invited to apply. Chris’s pitch: “Here’s a recommended list of candidates. Call these five people right now, and here’s why they’re a good fit. We’re saving recruiters time and helping them focus on the best candidates in their talent pools.” And it all lives right inside the job.

Once everything is sorted properly, we also show you what’s missing. It points out where a pool is thin, suggests categories you may not have thought to track, and makes it easy to fill the gap. You can post the right job, or bring in new candidates from three to four sources Boostie can search by occupation.

Travis sees all of this as a sourcing channel the industry has mostly ignored. Firms pay a lot for their ATS, he said, “and you don’t use it.”

6. Measure: proof it’s working

If you’re paying for software, you want to know it’s paying you back. That’s the first thing Travis wants an owner to see on the dashboard: time saved and money saved.

Time saved counts the work Boostie handled automatically. “Here’s how much time we’ve saved you,” Travis said. “You don’t have to do these rote tasks that you’re probably not going to do at all.”

Money saved starts from one idea: “Every applicant that you don’t buy from a job board has a value.” Boostie estimates that value using cost-per-applicant benchmarks from industry sources like the ASA, Bullhorn’s GRID report and SHRM.

Below that, the dashboard shows the health of your applicant funnel and how each source performs. Travis described an unhealthy funnel: 100 applicants, with 80% of them weak or very weak fits. “All the signals are pointing down.”

That’s where source quality matters. For about 30 years, Travis said, firms have spent on job boards and “just let it rip.” Now “you can be a little more discerning.” His bar for a good source: if 30 of 100 applicants are worth submitting to a client, that’s pretty good.

Soon we’re adding a coverage view, showing every open role and how many more applicants each one needs to reach a placement.

7. Flappy: the side project that went live

One day, Kevin was exhausted of new clients coming on board and having to answer the same questions again and again. So he built Flappy, Boostie’s AI team member. “I spent maybe half a day, maybe a day, and built something that integrated into our Slack. It worked way better than I thought it would. It was answering better than I would.”

Why an assistant instead of better help docs? Kevin was honest: “People just don’t look at the documentation. Or at least I don’t. I see a big paragraph, I’m like, I don’t want to read all that.” As a recently graduated college student, YES.

Flappy is now live in the Boostie app for every user. Ask it how something works and it answers. If it can’t, it hands you to the team. It’s also kept away from sensitive areas like fraud detection and security, so it can’t give away how those work.

Building it in-house was deliberate. “If we own the whole chatbot, we could do a lot more than just answer questions,” Kevin said. Down the road, that could mean Flappy noticing a job with only two applicants and offering to help.

Brian, who moved Flappy from Slack into the app, described the team’s approach: “Get it out there, get it in their hands, see what works, and see what the clients actually want and need as opposed to what I assume they want and need.”

Confession: Flappy also fact-checks everything I put out, including this.

Additional Note: While drafting, Chris wanted me to make it abundantly clear to everyone that Flappy can also handle weather reports. I have very strange bosses.

Tying it all together

None of this asks recruiters to learn something new. That was on purpose. “We don’t want you to have to use a prompt to be able to do anything on our platform,” Travis said. “We think automation should be a click, not a blank canvas you have to go build.”

Chris adds: “We don’t want to build another tool that takes six months to deploy. We want to deliver a digital workforce that can be up and running in 30 minutes to an hour, without requiring users to learn, adopt, or manage another tool. That’s how we deliver immediate value.”

Travis’s measure of success is that recruiters barely feel the change. The tedious work runs in the background, “so the humans can worry about doing human stuff and having real conversations with real people.”

Already a Boostie client? You may have more of this turned on than you think. Ask us.

Not a client yet? Book a demo, and we’ll show you what you’ve been renting.

Anyways, I’m gonna go talk to Flappy about the weather or something.