Expert in AI detection technology, Mykhailo Petrenko, explains how a surge in artificial resume content threatens hiring quality and what solutions can restore authenticity to recruitment processes
Nearly 70% of companies plan to incorporate AI into recruitment processes without human oversight by 2025, according to a recent Resume Builder survey of business leaders. This automation surge coincides with an unprecedented problem: AI-generated resumes that slip past traditional screening tools, creating what industry experts call an authentication crisis in hiring.
Mykhailo Petrenko witnessed this crisis firsthand. As an AI Agents Engineer at Apple, he was constantly contacted by recruiters reviewing applications that weren’t always what they seemed. His technical background research on neural network quantization published at AAAI 2023 and IEEE 2023 conferences, winning Google’s sustainability hackathon for AI innovation gave him unique insight into both how language models generate convincing content and how to detect it.
His solution, HireSight, has achieved 300% growth in B2B monthly active users, with early adopters reporting 17% reductions in screening time and 5% improved precision in candidate evaluation. But the real story isn’t just about detection technology. It’s about an arms race between content generation and authentication that will reshape how companies evaluate talent.
In this exclusive interview, Petrenko reveals why keyword-based screening systems are obsolete against modern AI, and why recruitment authentication will become as critical as cybersecurity in financial services. His experience from SoftServe in Poland to Apple in Austin offers a global perspective on a problem that transcends borders: in a world where both recruiters and applicants use AI, how do we preserve authentic human evaluation?
Mykhailo, AI-generated content in resumes has become a significant challenge for recruitment teams. How widespread is this authenticity problem, and what makes it so difficult to detect?
Current market conditions create perfect storms for this issue – AI tools make generating polished resumes trivially easy at scale, while companies receive record application volumes. Recruiters spend countless hours reviewing content that isn’t authentic candidate work, which fundamentally undermines hiring decisions. Volume pressure means traditional screening methods can’t distinguish between genuine experience descriptions and AI-generated text, but deep knowledge of neural network behavior enables more effective identification strategies. My research on neural network quantization, published in AAAI 2023 and IEEE 2023, revealed fundamental patterns in how language models generate text versus authentic human writing, and understanding these technical mechanisms enables more sophisticated detection approaches.
Given these technical challenges, how are new detection systems different from traditional ATS approaches that relied primarily on keyword matching?
Keyword-based systems treat resumes as single sources of truth without verification against other data points. Modern approaches integrate multiple information sources: social media profiles, organization verification, and career pattern analysis – to build comprehensive candidate pictures. Multi-source verification reveals inconsistencies that traditional systems miss entirely. Advanced scoring methodologies evaluate career trajectories, identify potential gaps or burnout indicators, and flag both concerning and positive signals. Context understanding replaces simple pattern matching, providing explanations for scoring decisions rather than black-box outputs.
Recognition from industry competitions like Google’s sustainability hackathon suggests growing awareness of AI applications in practical problems. What results are you seeing with your own solution in the market?
Winning Google’s sustainability hackathon in November 2024 for best use of AI taught me that practical applications require measurable outcomes, not just technical sophistication. With HireSight, I’m seeing 300% growth in B2B monthly active users, which indicates strong market demand for authentication solutions. Early adopters report 17% reductions in screening time per candidate and 5% improved precision by reducing false positives. More significant changes involve workflow quality. Recruiters tell me they feel much more confident when comprehensive candidate analysis replaces superficial resume reviews. Success comes from addressing real pain points rather than building technology for its own sake, similar to how sustainability solutions gain traction through demonstrated value.
International experience often provides unique perspectives on technology problems. How has your path from Poland to SoftServe, then Apple, and now entrepreneurship in Austin influenced your understanding of global recruitment challenges?
Working across different markets reveals how authentication problems scale globally but manifest differently. At SoftServe in Poland, I worked on ATS systems and saw how traditional keyword matching failed across languages and cultural contexts. Moving to Austin and joining Apple’s AI agents team exposed me to enterprise-scale challenges where the volume and sophistication of fake applications create unprecedented problems. Austin’s entrepreneurial community at Capital Factory taught me that successful solutions need global applicability but local market understanding. My international background helps me build systems that work across different hiring cultures while addressing universal authenticity concerns that transcend geographic boundaries.
Building successful AI applications requires both technical expertise and market understanding. You mentioned an earlier project that didn’t succeed. How did that experience shape your approach to solving recruitment authentication challenges?
My earlier feng shui layout webapp taught me invaluable lessons about market validation versus technical capability. That project failed because I built something I thought people needed rather than something I knew they needed from direct experience. With HireSight, I started from personal pain points. Being constantly contacted by recruiters while at Apple made me realize how broken the authentication process was. Earlier failure taught me that impostor syndrome and lack of marketing experience can block progress, but understanding that failure provides incomparable learning accelerated my development approach. Now I focus on solving problems I’ve personally experienced rather than theoretical market opportunities.
These results suggest detection technology is maturing rapidly. Where do you see the recruitment authentication space heading as AI capabilities continue advancing?
Arms races between content generation and detection will intensify, requiring constant algorithm updates and new verification methods. Transparency and explainability will become competitive advantages as regulatory scrutiny increases around AI hiring practices. Human-AI collaboration models will mature, with technology handling initial filtering and humans focusing on cultural fit, communication assessment, and complex decision-making. Authentication and verification infrastructure will become as critical to recruitment as cybersecurity is to financial services today. Market timing favors solutions addressing current pain points. AI agent growth intersects with record volumes of synthetic content.
Looking at your current trajectory, what are your plans for expanding authentication technology beyond resume detection?
HireSight currently focuses on resume authenticity, but broader verification challenges exist across the entire candidate journey. I’m exploring applications in interview authenticity – detecting when candidates use AI assistance during video interviews or assessment completion. Social media verification represents another frontier where employment history claims can be cross-referenced against actual professional networks and activities. Long-term vision involves creating comprehensive candidate authenticity platforms that serve as trusted verification layers for the entire recruitment ecosystem. My goal is to establish the same level of authentication infrastructure for talent acquisition that financial services use for transaction verification.
