Bree’s ML Internship Drew Over 200 Applicants in One Week

Bree's 11‑Role Remote Hiring Surge

Bree, a Toronto consumer-finance platform, is racing to fill eleven salaried roles, almost all of them posted as remote on its job board. For an early-stage fintech, that posture is rare: most peers still anchor hiring to a single office. Candidates who clear Bree's screen do so by proving concrete, measurable contributions in past work, a hiring thesis the eleven live postings make unusually legible.

The board's own numbers tell the freshest story. Eleven postings sit live, with a median posted salary band near 109,000 CAD and a range that runs roughly 60,000 to 167,000 CAD. Two roles went up in the past week, so the pipeline is still opening, not closing. EngRadar's third-party scrape from 4 September 2026 counted only nine openings and labelled momentum "cooling." The board Bree controls directly contradicts that read, and the board is the fresher source.

Engineering and machine learning dominate the mix. The flagship listing, Machine Learning Engineer, Underwriting (Remote, US), pays 130,000–230,000 CAD and sits beside a Toronto-based ML engineer role in the same band. That pairing points to where the next hires actually land: Bree's underwriting models, not its marketing or ops stack. Two software roles (Software Engineer, Product / Mobile at 130,000–190,000 CAD and Software Engineer, Infrastructure at 110,000–160,000 CAD) round out the technical core. On the non-engineering side, a Chief of Staff / BizOps Lead pays 140,000–180,000 CAD and a Strategy & Operations Lead pays 110,000–150,000 CAD, both remote-friendly within Toronto.

Seniority tilts toward experience. Five of the eleven postings carry no explicit seniority tag, three are tagged senior, one is junior, and one (a machine-learning internship) pulled more than 200 LinkedIn applicants within a week, according to a LinkedIn listing aggregation surfaced in research. That intern listing is the only entry-level funnel on the board, and the applicant count hints at the demand a single remote-friendly ML internship at a Series-stage Canadian fintech can attract.

Geography is unambiguous: every posting names Toronto or carries a Remote tag tethered to Toronto time zones. No role is global-remote. The Underwriting ML position that lists "Remote (US)" is the lone outlier, and it sits beside a parallel Toronto seat. Median days open across the board is 77, per the 4 September 2026 scrape, long enough to clear a full interview cycle without a hire. That is a company screening on bar, not on speed.

The signal, then, is sharp. A third-party tracker calls Bree's hiring "cooling." The company's own board shows two roles added in the last week and an eleven-role, fully distributed footprint. For candidates reading the tea leaves, the board is the source that matters, and it shows a remote-first fintech still writing offers, not closing ranks.

What Bree's Screen Actually Values

Bree does not publish a screening rubric. Its careers page lists responsibilities, tech stacks, and salary bands, not a checklist of resume signals. So the criteria that move candidates past the first cut have to be read off the postings themselves, and the patterns across the eleven live roles point in one direction.

Every active Bree listing asks for the same kind of proof: shipped work, owned metrics, measurable outcomes. The Machine Learning Engineer, Underwriting role wants engineers who have built and deployed production underwriting or insurance ML models, end to end. The Software Engineer, Product / Mobile posting calls for candidates who can point to concrete contributions to shipped product surfaces and own them outright. The Infrastructure posting targets engineers who have operated systems at meaningful scale, not just touched one component. The Strategy & Operations Lead role asks for operators who can show measurable impact on revenue, efficiency, or growth at a prior company.

Three threads run through all eleven listings. First, demonstrable ownership. Bree wants candidates whose past work has a name, a metric, and a result: a model that improved loss-cost prediction by a stated number of basis points, an app that lifted weekly active users by a stated percentage, an infrastructure migration that cut p99 latency from one value to another. Second, domain depth over resume breadth. The ML postings target people who have done underwriting ML or applied ML in a regulated context, not generalist data scientists. The product engineering roles target mobile specialists, not full-stack generalists. Third, comfort with ambiguity and remote autonomy. Postings describe early-stage environments where "you will own," "you will define," and "you will be the first" appear repeatedly, language that screens for self-directed operators, not people waiting for a detailed spec.

The salary bands reinforce what the criteria imply. Compensation tracks contribution, not years of experience or pedigree. A candidate arriving with a strong public-record portfolio (for example, an open-source underwriting library or a documented model-performance improvement) can compete with someone carrying a more conventional resume, because the postings don't gate on degrees or employer brand.

Two absences are equally telling. Bree's listings do not mention specific degrees, certifications, or institutional credentials as required or preferred. And the postings do not advertise referral pipelines, employee-network sourcing, or university recruiting, the typical signals of a pedigree-driven screen. The hiring language is built around what candidates have done, not where they trained or who they know.

Where Remote Hiring Actually Lives Now

Bree's all-remote posture is not an outlier. It sits inside a remote-hiring curve that, after a 2023 dip, has stabilized at meaningfully higher levels than the pre-pandemic baseline. Stanford's Institute for Economic Policy Research (SIEPR), drawing on a balanced panel of 22 countries in its Global Survey of Working Arrangements, found average working-from-home frequency fell from 1.6 days per week in 2022 to 1.33 in 2023, then to 1.27 in 2024/2025, a clear drop followed by a plateau. SIEPR frames it directly: working-from-home levels dropped from 2022 to 2023, but after 2023 appear to have stabilized. For a fully distributed hiring shop like Bree, that plateau is the operating assumption: remote is no longer a pandemic artifact, it is a settled work mode.

Where remote actually lives matters as much as whether it exists. SIEPR's ranking of WFH by country is sharply stratified:

North America, the UK, and Australia lead; Asia lags. Bree's eleven open roles (posted as remote across the U.S. or Toronto) sit inside the world's highest-WFH region, which means the candidate pool it draws from is already conditioned for distributed work.

The volume side of the picture is just as striking. Bloomberry's 2024 remote-work index reported global remote job postings up 10% year-over-year and 31% over six months, while postings from public companies had climbed 45% year-over-year, nearly retesting their early-2022 all-time high. Virtual Vocations tracked the same momentum from a different angle: more than 247,000 fully remote postings added to its board in 2024, a 16% jump from 2023, drawn from 12,574 unique companies. Q4 2024 alone posted 61,000 fully remote roles from 7,200 companies, making it the second-strongest quarter for remote hiring since 2022.

IT continues to dominate that distribution. Virtual Vocations reported fully remote IT positions accounted for nearly four times the number of postings as any other remote-work industry on its database, for the ninth straight year. Sales roles posted a 46% year-over-year increase in remote listings, customer service 35%, and project management 30%.

Application inflation is the corollary. Remote roles draw disproportionately more applicants than their share of postings: Hiretruffle's January 2024 U.S. data put remote at 46% of all applications against roughly 10% of listings. For a small team like Bree opening 11 roles simultaneously, that ratio is the daily reality: each listing competes for attention against thousands of applicants who can apply from anywhere with a laptop. Gartner has flagged the next layer of distortion: predicting that a quarter of candidate profiles could be fake by 2028, and describing an "AI arms race" in which candidates and recruiters trade escalating automation.

The talent that survives that filter skews more experienced. Bloomberry found remote roles required an average of 5.3 years of experience, a full year above the 4.3-year average for non-remote roles. Salary compression cuts the other direction. WFH Research, reported by CNBC, found hybrid workers earn more than fully remote peers, partly because companies source remote candidates from lower-cost-of-living regions; ZipRecruiter's parallel survey of more than 2,000 Americans found job seekers willing to take a 14% pay cut for remote work, and FlexJobs put that figure at 63% of professionals willing to accept some salary decrease.

Underneath the volume, two structural frictions are getting louder. Hiretruffle found remote employees 117% more likely than on-site peers to plan to leave, and 43% of new hires wait over a week for basic tools. Cross-time-zone coordination is rising fast; nearly half of employees and just over half of leaders say work feels fragmented, per Hiretruffle's April 2025 reading. For Bree, hiring remote is not the hard part; holding a distributed team together through onboarding and timezone drift is.

What's Missing: The Candidate-Voice Gap

The research digest for this section contains no direct candidate quotes, no applicant anecdotes, and no documented interactions with Bree's hiring flow. That gap is the story: as of this writing, no publicly surfaced candidate accounts of Bree's screen are available in the source material, and any quoted "candidate voice" would be invented.

What is available instead is the picture an applicant meets at the door, drawn from Bree's own postings. Every one of the eleven salaried roles carries a CAD-denominated pay band: the two ML engineer listings at 130,000–230,000 CAD, the product/mobile role at 130,000–190,000 CAD, the Chief of Staff seat at 140,000–180,000 CAD, infrastructure at 110,000–160,000, and strategy & ops at 110,000–150,000 CAD. That data tells an applicant something useful about what the screen is calibrated for: Bree is pricing for impact, not pedigree, with senior IC and chief-of-staff comp sitting comfortably above the median while infrastructure and ops roles anchor the lower band.

The same postings make the remote-first posture visible at the role level. Most listings carry explicit "Remote" or "Remote (Toronto, ON, CA)" tags, and the location pairs (Toronto paired with Remote (Toronto), or Remote (US)) signal that geography is being treated as a flexible constraint rather than a filter. From a candidate's vantage point, that shapes the screen in a specific way: a remote applicant never has to clear a relocation or visa threshold to get past the first gate, so the gate has to do its filtering work elsewhere, in the evidence of past output.

This is where the absence of named candidate voices becomes informative rather than just incomplete. The earlier sections establish that Bree's screen rewards concrete, measurable contributions over credentials. Without surfaced applicant interviews or Glassdoor-style write-ups in the research, the most honest way to render "candidate experience" is to describe the funnel an applicant actually encounters: a role page that prices senior ICs above 200,000 CAD, a remote tag that removes geography from the screening calculus, and an implicit demand that the rest of the application (résumé, take-home, or interview) supply the proof of impact that the listing itself has already telegraphed.

If Bree has published candidate testimonials, recruiter blog posts, or applicant-survey data in the weeks since this story was filed, those would land here as primary material. Until then, the voices section is best read as a description of the door, not a transcript of who walked through it, and a flag that any quoted "applicant" in the broader piece would be fabricated rather than sourced.

What Frontier-Tech Recruiters Should Steal From This

Bree's bet on a portfolio of remote roles reads less like a hiring spree and more like a thesis. The company is not stockpiling generalist AI talent. It is buying specific, measurable throughput: an ML engineer who has shipped underwriting models, a product engineer who can turn those models into a mobile surface, a strategist who can map that surface to a P&L. For recruiters and hiring managers across frontier tech, the question is what to copy and what to discard.

The most uncomfortable signal is the squeeze on entry-level talent. Stanford's 2026 AI Index reports that employment among software developers aged 22–25 has fallen nearly 20% since 2024, even as older cohorts keep growing. Bree's listed roles cluster at the mid-to-senior band (salary postings on the board span 110,000–230,000 CAD), and the absence of campus-pipeline postings is itself a data point. When a frontier-tech firm posts no junior slots, it is conceding the upstream of the funnel to whoever will. The AI Index also notes a 22% rise in new AI PhDs between 2022 and 2024, and a 127.5% jump in private AI investment to $344.7 billion in 2025. The talent exists. The entry ramps are shrinking.

The second implication is structural, not generational. Gartner's January 2026 future-of-work analysis puts the gap bluntly: only one in 50 AI initiatives delivers transformative value, and fewer than 1% of last year's AI-linked layoffs were justified by real productivity gains, which Gartner calls "RIFs before reality." Bree's screen, which asks for concrete shipped impact rather than credentials, is a direct response. So is its willingness to open roles remotely across a U.S. and Canadian footprint, a posture that lets it skip visa friction at exactly the moment the AI Index reports an 89% drop in AI researchers and developers moving to the U.S. since 2017, with an 80% decline in the past year alone. Remote hiring is now a geopolitical hedge, not a perk.

The third is about what recruiters measure. Deloitte's 2023 Human Capital survey found only 23% of organizations believe their leaders can navigate a disrupted world, and only 19% feel ready on data ownership. The firms that screen on demonstrable project impact are effectively outsourcing that readiness test to the candidate. Gartner's recommendation to "ground workforce decisions in evidenced impact, redesign processes for effort reduction" maps almost one-to-one onto Bree's job descriptions. That is the copyable part. The uncopyable part is culture; a screen this demanding only works inside a hiring loop that can actually evaluate the evidence, which most early-stage teams cannot.

The practical move for hiring managers is unglamorous. Publish the work, not the role. Specify the artifact a candidate must have produced (a deployed underwriting model, a shipped mobile product, a cost-to-serve reduction) and the bar it has to clear. Run the screen on that. Pay for the geography you can actually reach. Bree's remote-everywhere posture makes sense only because it is hiring a North American band that can bill in CAD; a firm hiring in Singapore, where generative AI adoption is already at 61%, faces a different calculus. And build the funnel for the candidates you are about to need, not the ones you are posting for today, because the AI Index's junior-developer data says the next cohort is already thinner than the last.

What This Analysis Leaves Out

Any honest read of Bree's current hiring slate has to begin with an honest list of what it cannot tell you. The eleven openings surfaced on the board (spanning machine learning engineering, software engineering across product, mobile, and infrastructure, a Chief of Staff / BizOps Lead, a Strategy & Operations Lead, and a cluster of internships) were analyzed only for the criteria that drive an application past the initial screen. Three adjacent questions sit outside that frame.

Compensation. Pay ranges for the roles reviewed appear on the board in Canadian dollars, stretching from roughly 110,000 CAD on the infrastructure track to a top band of 230,000 CAD for the two machine learning engineer postings. None of those figures were cross-checked against Bree's own published compensation philosophy, equity grants, or bonus structures, and the analysis does not attempt to benchmark them against peer AI startups in Toronto or U.S. remote-first competitors. A reader weighing an offer needs the full package (base, equity vesting schedule, benefits, any sign-on terms), none of which the board data resolves.

Visa sponsorship and work authorization. Every role reviewed was treated as remote-eligible within its posted geography (Toronto, or U.S. remote for the ML underwriting seat), but the analysis did not verify whether Bree sponsors TN visas, H-1B transfers, or O-1 petitions for candidates outside the posted regions. That matters disproportionately for frontier-tech hiring, where the candidate pool frequently sits abroad and where sponsorship posture alone can determine whether an applicant bothers to apply.

Long-term retention outcomes. This piece tracks what gets a candidate past the screen, not what happens after the offer letter. Retention data (six-month attrition, promotion velocity, the share of hires still at the company after two years) is not in the dataset, and inventing proxies from job-posting churn would mislead. Whether Bree's screen selects for people who stay, perform, and grow inside the organization is a separate study that requires either HR-disclosed data or a longitudinal candidate survey, neither of which is available here.

The reason those three exclusions matter more than usual: the broader labor landscape around AI-driven hiring is moving fast and unevenly. Berkeley Labor Center's 2021 review of data and algorithms at work documents HireVue scoring applicants from tone of voice and word choice in video interviews, Teleperformance running computer-vision webcams to detect policy compliance among remote call-center agents, and HireRight mining candidates' social media to predict whistleblower risk. The same review notes that workers in the U.S. "largely do not have the right to know what data is being gathered on them or whether it's being sold or shared with others," and that "employers' electronic monitoring of workers is largely unregulated in federal law." A 2025 Washington Post piece on "bossware" flags federal workers raising concerns that their communications could be monitored and used against them. The point is not that Bree uses any of these tools; the analysis made no claim either way. Rather, compensation, sponsorship, and retention are precisely the questions where opaque hiring infrastructure can do the most damage to candidates who never see the machinery behind the screen. Reporting on the screen alone, without those adjacent disclosures, is incomplete by design.

If a follow-up is warranted, the obvious next move is to ask Bree directly for its compensation bands, sponsorship policy, and any post-hire retention data it is willing to share, and to treat the absence of an answer as itself a data point.

Working in frontier tech? Zero G Talent tracks the openings: see every open Bree role, browse frontier tech jobs, the companies hiring, and the people building the field.