The hiring funnel has a new, AI-shaped hole. After a resume is parsed and before a first interview is scheduled, there’s a growing question of what a candidate can actually do with the tools now on every desk. Oppi AI, a Helsinki-based applied-AI company, is betting the answer lies in a 45-minute simulated workday.
Its product, WorkProbe, generates a realistic task from a job description, complete with source files, a simulated AI assistant, and workplace context. The goal isn’t to test rote skills, but to evaluate how a candidate reasons under uncertainty, detects inaccuracies, and,critically,how they interact with AI [oppi.ai]. It’s a wedge into the assessment layer of hiring, positioned for employers who suspect traditional screening is increasingly obsolete.
From Education to Enterprise Hiring
Oppi AI was founded in 2018, and its public footprint shows a pivot. Earlier descriptions positioned it as a deep-tech education platform using machine learning to personalize student learning [oppi.ai]. Today, the company’s website and LinkedIn profile are singularly focused on WorkProbe for hiring teams [LinkedIn, September 2026]. This shift from B2C education to B2B talent assessment is a significant strategic turn, suggesting the founders identified a more acute enterprise pain point in the AI era.
The technical founder credentials are research-heavy. CEO Taras Zagibalov holds a PhD in computational linguistics from the University of Sussex and has a background in NLP and data science, including early research at Brandwatch [oppi.ai]. A historical company profile lists Aleksandr Zaretckii as CTO, with dual PhDs in applied mathematics and information technology [Gust, November 2025]. The team’s depth in language and AI research is the foundation for a product that claims to measure nuanced human judgment.
The Wedge: Measuring AI-Era Soft Skills
WorkProbe’s differentiation is its focus on meta-skills. The company argues that in a world of AI co-pilots, the valuable employee isn’t the one who knows the most, but the one who can best direct, critique, and trust the output of AI tools. Its marketing highlights six evaluation dimensions: prompt quality, critical filtering, adaptation speed, AI awareness, calibrated trust, and judgment under ambiguity [oppi.ai].
The workflow is designed for procurement ease. Hiring managers reportedly generate a simulation from a job description with no integration required, slotting it between initial screening and interviews [f6s.com, September 2026]. The output is a ranked shortlist of candidates based on “skills a CV cannot show” [workprobe.com, 2026]. For a hiring manager overwhelmed by AI-polished resumes, the promise of a ready-to-use, judgment-focused filter is clear.
Traction and the Capital Question
Public traction metrics are limited, but directional. The company states it is “onboarding pilot companies now” on its LinkedIn profile [LinkedIn, September 2026]. Third-party databases report a small team and modest, inconsistent funding figures, which is typical for an early-stage European deep-tech startup navigating a pivot.
| Metric | Reported Figure | Source | Confidence |
|---|---|---|---|
| Total Employees | 2 | [PitchBook, 2025] | ORANGE |
| Total Funding (attributed) | $1.39 million | [CB Insights] | ORANGE |
| Latest Round (attributed) | $890,000 | [CB Insights] | ORANGE |
| Total Investors (attributed) | 2 | [Tracxn, 2025] | ORANGE |
These figures should be read as signals of activity, not definitive financials. They point to a company that is capitalized to build and pilot, not yet to scale a global sales motion. The patent-pending status cited for WorkProbe is a common step for defensibility, but it’s a process claim, not a granted monopoly [LinkedIn, September 2026][patents.stackexchange.com, retrieved 2026].
Where the Bet Gets Hard
The ambition is compelling, but the path to enterprise revenue is lined with specific hurdles. Oppi AI is not just selling a new feature; it’s asking HR departments to insert a novel, unproven step into a long-established and often rigid process. The risks are less about technology and more about procurement behavior.
- Process inertia. Inserting a 45-minute assessment pre-interview adds friction for candidates and requires buy-in from recruiters and hiring managers accustomed to their own workflows. The no-integration claim helps, but adoption requires changing human behavior.
- Benchmarking validity. The product’s entire value rests on the correlation between its simulation scores and on-the-job performance. Building that validated, industry-specific benchmark dataset is a long, expensive undertaking that precedes any claim of predictive power.
- The commoditization front. The core concept,AI-generated assessments,is not defensible by itself. The moat must be the proprietary scoring model and the quality of the simulation library. Without constant iteration, this is a space where larger HR tech platforms could quickly replicate a basic version.
The company’s answer likely hinges on pilot outcomes. If early design partners report dramatically better hire quality and provide public case studies, that evidence becomes the fuel for a land-and-expand motion. Without those concrete success stories, it remains an interesting tool in search of a budget owner.
The Next Twelve Months
For Oppi AI, the immediate future is about proof. The next milestones are not feature launches, but commercial validations. The key watchpoints are the conversion of those onboarding pilots into publicly named design partners, and the subsequent announcement of a priced enterprise contract. A seed round to fund the build-out of a commercial team and accelerate benchmark development would be a logical next step, given the attributed funding history suggests runway for building, not scaling.
The ideal customer profile here is a tech-forward, mid-size company (roughly 200-2000 employees) that hires for knowledge-worker roles where AI tool usage is already endemic. Think software engineers, product managers, marketing analysts, or data scientists. These organizations are large enough to feel hiring pain acutely but small enough to pilot new tools without enterprise procurement cycles. They are also the most likely to have hiring managers who immediately grasp the problem WorkProbe is trying to solve.
The realistic competitive set isn’t other simulation startups,it’s the assessment modules baked into existing Applicant Tracking Systems (ATS) like Greenhouse or Lever, and the entrenched behavioral assessment giants like SHL. Oppi AI’s bet is that those incumbents are too slow to build for the AI era, and that their product is sufficiently differentiated to be bought as a point solution, not as part of a suite. It’s a classic wedge play. The next year will show if the wedge is sharp enough.
Sources
- [oppi.ai, September 2026] Oppi AI company website | https://www.oppi.ai/
- [LinkedIn, September 2026] Oppi AI LinkedIn profile | https://www.linkedin.com/company/oppi-ai
- [f6s.com, September 2026] Oppi AI startup profile | https://www.f6s.com/company/oppiai
- [workprobe.com, 2026] WorkProbe product website | https://www.workprobe.com/
- [Gust, November 2025] Oppi AI Oy company profile | https://gust.com/companies/oppi-ai
- [CB Insights] Oppi AI funding profile | https://www.cbinsights.com/company/oppi
- [PitchBook, 2025] Oppi AI 2025 Company Profile | https://pitchbook.com/profiles/company/459184-87
- [Tracxn, 2025] Oppi AI company profile | https://tracxn.com/d/companies/oppi-ai/__XFBtU3db_rJopQy6MJOKkK7CyHCFsaAQuovL0Gpe4Qs
- [patents.stackexchange.com, retrieved 2026] Patent pending definition | https://patents.stackexchange.com/questions/20236/patent-pending-us-and-eu-procedure-and