Decatron.ai
Post-transformer Compute for Real-time Cognitive Intelligence
Company Overview
Snapshot
Founded in April 2026 by Miki Lumnitz and Gal Rotem, Decatron.ai is currently in stealth mode with 1–10 employees. The company is developing a novel post-transformer EBM compute architecture designed to enable real-time cognitive intelligence.
Business overview
Decatron.ai is a Business Software company focused on IT, R&D & Data Solutions, specifically in Software Development Tools. The company is developing a deterministic post-transformer EBM compute architecture, centered around its proprietary DecaCell, a 10D geometric compute unit. This technology aims to replace traditional AI stacks that rely on massive models and dense tensor computation. Decatron.ai's architecture powers GERVIS, its native AGI foundation, and DecaQ, designed for Quantum-Class Computation without requiring a Quantum Processing Unit. It serves enterprise and professional services, particularly in the IT sector, by enabling lawful cognitive intelligence for real-time operational decision-making.
Log in to access full profile ›Company Intelligence Q&A
- What is Decatron.ai's core technological innovation?
- Decatron.ai is developing a post-transformer EBM compute architecture, built around its proprietary DecaCell, a 10D geometric compute unit. This architecture aims to provide deterministic computation for real-time cognitive intelligence, moving beyond traditional AI's reliance on probabilistic prediction and dense tensor computation.
- When was Decatron.ai founded?
- Decatron.ai was founded in April 2026.
- Who are the founders of Decatron.ai?
- Decatron.ai was co-founded by Miki Lumnitz, who serves as CEO, and Gal Rotem.
- What products does Decatron.ai's technology power?
- The DecaCell EBM compute architecture powers GERVIS, Decatron.ai's native AGI foundation, and DecaQ, which is designed for Quantum-Class Computation without a Quantum Processing Unit.
- What problem does Decatron.ai aim to solve in the AI landscape?
- Decatron.ai aims to address the limitations of current AI, which depends on massive models, probabilistic prediction, and heavy data movement, by introducing a deterministic compute architecture for real-time cognitive intelligence in operational domains.