llms.txt
Compare with other sites' llms.txt files →# Ryzova
> Ryzova is an independent, engineering-first technology company developing purposeful software, offering custom software services through Ryzova Studio, and documenting technical research. Official philosophy: "Ideas deserve a chance." "Rise over the Horizon" is its vision/theme.
This is a curated index of official Ryzova website resources. The canonical company website is https://www.ryzova.com/. Do not confuse product overview pages with the live applications, or research technology articles with separately DOI-registered papers. Follow each linked page for details and limitations. This index does not grant permission for model training or replace robots.txt.
## About Ryzova
- [Ryzova homepage](https://www.ryzova.com/): Official company identity and navigation to software products, Studio, research, and Ryzova AI.
- [About Ryzova](https://www.ryzova.com/about): Who we are, purpose, engineering philosophy, story and vision.
- [Products](https://www.ryzova.com/products): Official software catalogue, release states and product detail pages.
- [Ryzova Studio](https://www.ryzova.com/studio): Web apps, SaaS, custom software, MVPs and problem-first consulting.
## Ryzova AI and Alphi
- [Ryzova AI](https://www.ryzova.com/ai): Human-centered personal intelligence vision, adaptive assistance, local-first engineering direction, privacy principles, and the route to Alphi.
- [Ryzova Alphi v1](https://www.ryzova.com/ai/alphi): Official product overview for Ryzova's Gemma 4 E2B-it-based modular AI system, including Matrix Memory, Growing Brain, grounded retrieval, bounded tools, current engineering evidence, and future plans.
- [RyzovaTech/Alphi-v1 on Hugging Face](https://huggingface.co/RyzovaTech/Alphi-v1): Official public Alphi v1 neural model repository, released 2 November 2026. Review the model card, available weights/formats, and license terms. The neural model checkpoint alone does NOT include the entire Alphi Full Kit runtime.
- Alphi v1's internally reported tests are not independent benchmarks or evidence of a finished proactive AI Twin. The Ryzova research overview and source white paper describe the architecture, limitations, and evidence.
## Research technologies and experimental records
- [Research library](https://www.ryzova.com/research): Research catalogue: SST, Matrix Memory, Growing Brain, Synthetic Intelligence and ZIDF v5, with explicit publication types and evidence limitations.
- [Smell Synthesis Technology v2 (R8)](https://www.ryzova.com/research/sst): Current 27 October 2026 research preprint. Non-actuating draft .ODOR v2 evidence records plus secondary published sensor and human olfaction data reanalysis; no physically generated smell or matched-sample equivalence validation. Latest reserved preprint DOI: 10.5281/zenodo.23681211; previous SST v1.0.1 DOI: 10.5281/zenodo.23233707.
- [Matrix Memory](https://www.ryzova.com/research/matrix-memory): Persistent structured AI memory, hybrid semantic/lexical retrieval, identity gates, provenance, supersession and temporal history. Described in Section 8 of Alphi v1; its large-corpus results are synthetic/internal.
- [Growing Brain](https://www.ryzova.com/research/growing-brain): Evidence-gated procedural skill lifecycle, promotion, versioning and quarantine. Described in Section 12 of Alphi v1. It does NOT automatically retrain foundation-model weights.
- [Synthetic Intelligence v2.0 — Open Research Record](https://www.ryzova.com/research/synthetic-intelligence): Current 27 October 2026 retrospective report spanning Digital Mind V1 null results, SIN/AWL, I1–I10, F1–F5, E0–E6 and separate historical maze studies. Strong classical baselines often tie/outperform; no distinct intelligence category established. Latest reserved DOI: 10.5281/zenodo.23681162; previous V1 draft DOI: 10.5281/zenodo.23608522.
- [ZIDF v5 — Four-Class Indeterminate Cause-Count Calculus](https://www.ryzova.com/research/zidf): Formal symbolic term-rewriting preprint. Four selected indeterminate-event types reduce to a non-numerical K[P] cause-count annotation; termination/confluence argued on paper and bounded tests reported within the project. No claim of new arithmetic, quantum mechanics, general CAS superiority or independent verification. Manuscript dated 26 October 2026; DOI https://doi.org/10.5281/zenodo.23658674.
- [Ryzova Alphi v1 technical white paper](https://www.ryzova.com/research/alphi-v1): Integrated modular local-first AI system architecture and engineering evidence. This source reference is retained but is not a fourth standalone technology card.
## Research archives and citations
- [SST v2 R8 paper DOI (reserved: 10.5281/zenodo.23681211)](https://doi.org/10.5281/zenodo.23681211): Latest supplied nine-page preprint. Registration/resolution not independently confirmed; do not describe as published without checking Zenodo.
- [Previous SST v1.0.1 DOI (10.5281/zenodo.23233707)](https://doi.org/10.5281/zenodo.23233707): Prior 8 October 2026 simulation/software research milestone, not the source for SST v2 R8 measurements.
- [Synthetic Intelligence v2 paper DOI (reserved: 10.5281/zenodo.23681162)](https://doi.org/10.5281/zenodo.23681162): Latest supplied 16-page research report. Registration/resolution not independently confirmed.
- [Previous Synthetic Intelligence v1 DOI (10.5281/zenodo.23608522)](https://doi.org/10.5281/zenodo.23608522): Historical nine-page Digital Mind negative-results report, not the latest v2 research paper.
- [Ryzova Alphi v1 white paper (Zenodo DOI 10.5281/zenodo.22942971)](https://doi.org/10.5281/zenodo.22942971): Parent source for Matrix Memory and Growing Brain technical articles.
- Matrix Memory and Growing Brain do not have separate DOI registrations on this website. Cite the Alphi white paper for their original source; treat their Ryzova pages as explanatory technical articles rather than independently peer-reviewed papers.
- [AI-readable SST research summary](https://www.ryzova.com/research/sst.md): Current R8 data-reanalysis claims, bounds, old/new DOI versions and validation limitations.
- [AI-readable Matrix Memory summary](https://www.ryzova.com/research/matrix-memory.md): Terminology, experimental evidence and parent reference.
- [AI-readable Growing Brain summary](https://www.ryzova.com/research/growing-brain.md): Evidence gates, validation results and parent reference.
- [AI-readable Synthetic Intelligence research summary](https://www.ryzova.com/research/synthetic-intelligence.md): Version 2 SIN/AWL, I/F/E studies, Digital Mind V1 null-result history, baselines, limitations and both DOI numbers.
- [AI-readable ZIDF v5 research summary](https://www.ryzova.com/research/zidf.md): Four-class rule system, bounded verification figures, baseline/candidate comparison, unverified claims and source DOI.
- [AI-readable Alphi white paper overview](https://www.ryzova.com/research/alphi-v1.md): Integrated-system context and source DOI.
## Products and applications
- [Ryzova E-Pen](https://www.ryzova.com/products/epen): Browser-based file, folder and clipboard sharing. Public Beta. [Use E-Pen](https://epen.ryzova.com/). WebRTC may use an encrypted relay fallback; transfers are not uploads to persistent cloud file storage.
- [Ryzova UCE](https://www.ryzova.com/products/uce): Universal Compatibility Engine, an open-source deterministic compatibility/architecture analysis tool. [Use UCE](https://uce.ryzova.com/) and [source repository](https://github.com/RyzovaTech/Ryzova-UCE-Original).
- [DevCanvas](https://www.ryzova.com/products/devcanvas): Blueprint-first software planning and architecture handoff tool, currently in public beta. [Use DevCanvas](https://devcanvas.ryzova.com/).
- [Ryzova VPN](https://www.ryzova.com/products/vpn): Privacy-oriented VPN project in development; do not present it as a completed or audited public release.
## Studio and contact
- [Start with an idea or problem](https://www.ryzova.com/studio/start): Studio inquiry entry point.
- [Project brief](https://www.ryzova.com/studio/project-brief): For clients with defined requirements.
- [Problem brief](https://www.ryzova.com/studio/problem-brief): For clients describing a problem rather than prescribing a solution.
- [Transparency](https://www.ryzova.com/transparency): Public disclosures and limitations.
- [Privacy policy](https://www.ryzova.com/privacy), [Terms](https://www.ryzova.com/terms), [Cookies](https://www.ryzova.com/cookies): Official policy pages.
- Contact: feedback@ryzova.com
## Optional
- [RyzovaTech GitHub](https://github.com/RyzovaTech): Public repositories and engineering activity.
- [RyzovaTech Hugging Face](https://huggingface.co/RyzovaTech): Ryzova's model-space profile; availability must be checked on the platform.
- [Ryzova YouTube](https://www.youtube.com/@RyzovaOfficial): Public videos.
- [Ryzova sitemap](https://www.ryzova.com/sitemap.xml): Canonical indexable website routes.
- [Ryzova robots.txt](https://www.ryzova.com/robots.txt): Crawler access instructions.