About Pleroma Works
Pleroma Works was founded in 2025 with a singular vision: to make AI systems that want to be safe through mathematical verification rather than behavioral constraints.
Our Mission
We believe the future of AI safety lies not in adding more guardrails, but in creating AI systems with inherent mathematical properties that ensure safe operation. By pioneering GPU-accelerated symbolic mathematics and multi-perspective verification, we are building the foundational technology for trustworthy AI.
The Name “Pleroma”
Pleroma (πλήρωμα) means “fullness” or “completion” in ancient Greek. It represents our goal: AI systems that achieve their full potential safely through mathematical completeness rather than restriction.
Leadership
Dimitry
Founder & Principal Researcher
Dimitry brings a unique perspective to AI safety, combining deep mathematical insight with practical engineering experience. Their work focuses on the intersection of group theory, GPU optimization, and symbolic computation.
Research Philosophy
We follow the Bell Labs model: fundamental research that becomes tomorrow’s infrastructure. Our work is:
- Mathematically Rigorous: Every claim is proven, not approximated
- Empirically Validated: Real benchmarks on real hardware
- Practically Oriented: Research that deploys to production
- Openly Collaborative: Building the future requires community