When we built MindEscape, our goal was clear: replace static personality questionnaires with immersive escape-room experiences that reveal how people actually think, plan, and behave under pressure. What we didn’t anticipate was how quickly the platform would move beyond HR and recruitment into a completely different domain—venture capital and startup sustainability.
Today, MindEscape is a core component in a growing body of academic research that uses AI to predict which startups are most likely to succeed. Four peer-reviewed publications—spanning top journals and international conferences—now document how game-based personality assessment can improve investment decision-making.
This work comes from a collaboration between Dr. Nikolaos Takas (Senior Partner & CEO of THERMI Venture Capital Management), our SeriousMind team, and researchers from the Department of Informatics and the Department of Economics at the Aristotle University of Thessaloniki, under the supervision of Professor Ioannis Vlahavas and Professor Dimitrios Kousenidis.
The central idea is simple but powerful: experienced investors often say that founders’ traits—calmness under pressure, organization, execution focus—matter as much as the product or market. Until now, that intuition was hard to measure systematically. MindEscape changes that.
1. AI Based Founders’ Personality Assessment as Key Factor for Startup Sustainability
N. Takas, G. Liapis, K. Moutsianas, E. Kouloumpris, I. Vlahavas, D. Kousenidis
Technology, Knowledge and Learning, Springer (forthcoming)
This paper shows how MindEscape’s behavioral data replaces traditional Big Five questionnaires in a machine learning model for startup viability. Instead of asking founders how they describe themselves, the game records how they actually behave when solving puzzles under pressure. The result: founders with low neuroticism and high conscientiousness—calm, organized, execution-focused—show measurably higher chances of success.
2. Startup Sustainability Forecasting through Natural Language Processing of Business Plans
N. Takas, E. Kouloumpris, K. Moutsianas, G. Liapis, I. Vlahavas, D. Kousenidis
Submitted to Journal of Global Entrepreneurship Research, Springer
This work focuses on extracting predictive signals from business plans using natural language processing. Combined with other data sources, NLP features contribute to a multi-modal AI system that evaluates startups across three dimensions: market desirability, technological feasibility, and financial viability.
3. Startup Sustainability Forecasting with Artificial Intelligence
N. Takas, E. Kouloumpris, K. Moutsianas, G. Liapis, I. Vlahavas, D. Kousenidis (2024)
Applied Sciences, 14(19), 8925
https://doi.org/10.3390/app14198925
Published in Applied Sciences, this paper presents the complete AI architecture for startup viability prediction. Experiments with seven machine learning algorithms on real Greek startup data showed that ensemble methods—particularly Random Forest—achieved the highest accuracy. Critical variables identified include profitability, MVP existence, and annual revenue above €57,000.
4. Multi-Modal AI Startup Sustainability Forecasting System with Game-Based Personality Assessment
N. Takas, G. Liapis, E. Kouloumpris, D. Anagnostopoulos, I. Vlahavas, D. Kousenidis
Proceedings of the 22nd International Conference on Artificial Intelligence Applications and Innovations (AIAI 2026), Chania, Greece
This conference paper brings together all components of the system: business plan NLP, digital footprint analysis, financial indicators, and MindEscape’s personality assessment. The research confirms that incorporating founder personality data reduces prediction variability by 36%, making the system more stable and reliable for investment decision-making.
These publications establish something important: MindEscape is not just a recruitment tool. It’s a research-validated platform for understanding human behavior in ways that matter for real business outcomes—from hiring decisions to investment decisions.
For SeriousMind, this body of work validates years of investment in combining psychology, neuroscience, game design, and machine learning. It also shows how academic research from AUTH can produce tools with tangible impact, from the lab to the market.
And for the broader AI and entrepreneurship ecosystem, it offers a new model: using serious games to capture behavioral data that’s richer, more objective, and more predictive than traditional self-reports.
MindEscape started as an escape room. Today, it’s part of the scientific infrastructure that helps investors, accelerators, and policymakers make smarter decisions about which startups deserve support—and which founders have what it takes to build something lasting.
© 2026 SeriousMind. All Rights Reserved. Privacy Policy