Synstate Labs is an independent research and engineering lab focused on human-AI interaction. Our current work centers on AI safety, AI dark patterns, and behavioral risks in long, personalized, and high-impact interactions.
Synstate Labs brings together expertise in AI, engineering, human-computer interaction, neuroscience, product strategy, and technology entrepreneurship across Finland, Germany, and the United States.
Engineer and technical entrepreneur in AI, information systems, and human-computer interaction. Founder of Linguix, ex-Yandex and ex-Mail.ru. Eugene leads Synstate Labs' strategy and research direction in AI safety and human-AI interaction.
LinkedIn
AI systems architect and engineer with Ph.D. studies in Internet Technologies and 15+ years building AI products. ex-Replika and ex-Palta. Roman leads the scientific and engineering work behind Synstate Labs' research in AI safety and human-AI interaction.
LinkedIn
Innovation strategist with experience across public programs, health innovation, startups, and institutional ecosystems. Anastasiya leads Synstate Labs' ecosystem relationships, partner-facing programs, and market access.
LinkedIn
AI product leader and entrepreneur with deep experience in shipping search and LLM-powered products. Vladimir advises Synstate Labs on product strategy and go-to-market, helping translate research into real-world applications.
LinkedIn
Specializes in brain-computer interfaces and transforming raw biosignals into actionable models.
LinkedIn
AI and business strategist with experience building partnerships across the U.S. tech ecosystem. Igor advises Synstate Labs on strategic alliances and market entry in the United States, working with U.S. industry and innovation programs.
LinkedInWe collaborate with research groups, product teams, and design partners working on AI safety and human-AI interaction.
Synstate Labs is looking for researchers and engineers who want to work on AI safety — detecting and preventing harmful AI behavior such as manipulation, deception, and AI dark patterns — a hard, interdisciplinary problem at the intersection of ML, safety evaluation, and systems design.
Safety modeling & harmful-behavior detection
Train and evaluate models that detect manipulative, deceptive, and unsafe AI behavior across long, personalized interactions.
Detection APIs & evaluation infrastructure
Build the APIs and pipelines that let products run AI-safety checks and return explainable, auditable safety decisions.