From getting in the door to growing once you’re through it, these are the companies using technology to build people up, not push them out
AI hasn’t touched every part of work equally. One of the places it’s hit hardest is screening, hiring, and developing a team; human resources tasks that touch essentially every industry. A recent survey of business leaders found 82% of companies already use AI to screen resumes, 64% use it to evaluate candidate assessments, and close to a quarter use it to run interviews. It’s also where most of the fear lives. AI has limited entry-level roles and has made the job-hunting process that much harder to crack.
The irony is that the same technology causing that problem is the one a group of leaders is using to fix it. Some are pointing AI at scaling connection and skills to develop the people still needed to do a job only a human can do. Others are pointing it at replacing a workforce, chasing a quick save on the bottom line. Same tool, two very different approaches on where the value lies.
The companies doing this well aren’t just asking what AI can save them this quarter. They’re thinking about where this is headed and what it takes to build something that actually holds up,
The ten companies below are on that side of the line. While many companies hop on buzzwords like “human in the loop”, these companies are actively prioritizing the human within it all.
Remoti: Rebuilding the workforce as infrastructure
Remoti treats hiring as infrastructure, not a one-off transaction. Its Workforce-as-a-Service model bundles recruiting, compliant hiring, payroll, IT, and retention into one operating layer, so a company can stand up and run entire teams across Latin America. The hardest part of global hiring is usually everything that happens after the offer letter goes out, and that’s what Remoti is actually built to own. It isn’t just matching candidates to a role. Remoti is the employer of record, doing the hiring itself while the client company directs the work.
Build Talent Labs: Widening the door for global talent
The hardest barrier to work often isn’t skill, it’s paperwork. Build Talent Labs runs structured immigration programs, cap-exempt H-1B, O-1 for people with extraordinary ability, J-1 for trainees and scholars, that let U.S. companies hire and keep global talent without waiting on a lottery. It isn’t advising from the sidelines either. Build Talent Labs is usually the one filing as sponsor and making the hire directly. A future-of-work story built mostly around algorithms could use a reminder like this one: getting someone to opportunity is still partly a human, legal, and logistical problem.
Gorilla Logic: Choosing people over headcount
Senior-first hiring isn’t new for Gorilla Logic; it predates AI entirely. What’s changed is the reasoning behind it. Its own engineers argue that AI raises the cost of a bad decision instead of lowering it. A flawed pattern used to land in one file. Now an agent can copy it across an entire codebase before anyone reviews a line of it. Having someone senior in the room who can catch that before it spreads matters more now, not less.
That doesn’t mean the junior path disappears, it’s just being reshaped. Instead of writing code from scratch, new engineers learn by reviewing and fixing what the AI produces, working on real systems, and explaining why a model got something wrong instead of just patching over it. It’s a company treating AI as something that raises the value of judgment rather than something that replaces the people still building it.
Kendo AI: Hiring and growing people with the same tool
Kendo actually sits on both sides of this list. Before a sales candidate gets an interview, a hiring manager can run them through an AI roleplay screen that tests objection handling, cold calling, and closing under the same scoring system Kendo uses on real reps, so a bad hire gets caught before anyone spends an hour with them. Once someone’s hired, that same system reviews their actual calls and coaches them against the company’s own methodology, with realistic AI prospects to practice against before the stakes are real. It’s a rare case of one company building for getting in the door and getting better once you’re through it.
Mercor: Prove yourself to a machine, then get paid
Mercor vets professionals, including engineers, doctors, and lawyers, through a roughly 20-minute AI video interview, then matches them to paid work, a lot of which involves training the same AI models reshaping their fields. It flips the old order on its head. Instead of proving your expertise by getting hired, you prove it to an algorithm that then sends opportunity your way. It complements Remoti rather than competing with it, since it’s a marketplace rather than managed teams, and it’s a vivid answer to the question of what new work AI actually creates.
Matcha: Giving the candidate the upper hand
Matcha goes straight at the anxiety, the spray-and-pray job hunt where 150 applications get you one offer. It matches seekers to roles on real fit instead of keyword overlap, and it shows candidates their match score and the reasons behind it before they apply, so they stop guessing. It’s free for job seekers, which tilts the power back toward the person trying to break in.
Eightfold AI: Finding the people already in the building
Eightfold’s deep-learning engine reads skills rather than job titles, so it can surface a current employee who could move into an open role they’d never be found for by keyword, or a candidate whose transferable skills a human screener would miss. As entry-level opportunities get scarcer, its internal-mobility angle points to a future where careers get built by moving within organizations, not only between them.
Cloverleaf: Keeping AI from becoming an exit counselor
Cloverleaf sits at an awkward intersection in this list: it is building AI for workplace coaching while its own research is showing where generic AI coaching can go wrong. A recent Cloverleaf Labs report tested five leading LLMs across five common workplace conflicts, producing 75 responses that were reviewed by trained human evaluators. The models generated 638 distinct pieces of advice, but only three encouraged employees to genuinely invest in repairing the relationship. In the two manager scenarios, 27 of 30 responses raised leaving the job as a legitimate option. The report also found relational coaching dropped by roughly 40% when the person asking for advice had less power. The point isn’t that AI has no place in workplace coaching. It’s that helping someone understand another person is a different job from helping them manage a problem, and current LLMs don’t consistently make that distinction.
Metaview: Turning interviews into intelligence
Metaview records and structures hiring conversations, turning a 30-minute screen into a consistent, reviewable candidate report within the hour. The idea is fairer, faster decisions with less bias, and less of the “rejected in four minutes by someone who clearly didn’t read it” experience that’s wearing down candidate trust.
Sonata Software: Reskilling the workforce for the AI era
Sonata Software approaches the AI shift less as a headcount question than a skills question. Its modernization business is increasingly built around AI, but the company is also putting that technology into its own workforce. In a February investor update, Sonata said more than 92% of its workforce and 80% of its managers had been trained in AI, while 78% of employees had completed live coding training. That matters because the transition to AI changes what experienced engineers are expected to do, not just how many people a company needs. Sonata’s model is to push the workforce toward new capabilities as the technology changes underneath it, making reskilling part of the operating model rather than a side program.
Disclosure: This article mentions clients of an Espacio portfolio company.
