VylexAIVylexAI
GitHub →
EU-compliant distributed AI compute network · regulated & supervised · Made in Germany

Every device,
a node of the global AI network.

VylexAI transforms everyday devices into a distributed network for AI compute and secure data storage. Millions of idle smartphones, tablets, and PCs — unified into one verifiable layer that trains models, serves inference, and holds encrypted data.

See the code →How it works
VylexAIDudaev Systems UG · Cottbus, Germany · Aligned with EU AI Act
How it works

Three steps, one network.

STEP 01/ 1

Install the app

Download VylexAI on Android. Create an account, or continue as a demo user to skip straight to the experience.

STEP 02/ 2

Your phone becomes a node

While charging on Wi-Fi at a safe temperature, the app runs small AI tasks in the background. You earn BSAI — the network's unit of account — for every verified task.

STEP 03/ 3

Or run your own AI

Submit inference or light fine-tuning jobs. The network dispatches them across participating devices and returns the results quickly, at a fraction of cloud cost.

Architecture

Four layers. One coherent system.

We don't improve the old AI stack — we replace it. Storage, compute, AI and verification woven into a single distributed fabric.

Storage

Distributed data storage

Encrypted, erasure-coded files distributed across independent nodes. No single point of failure, no cloud lock-in.

Compute

Global compute marketplace

GPUs, CPUs, and phones become a unified pool. Tasks are routed to whichever nodes are closest, cheapest, and most capable — every result verifiable in the coordinator ledger.

AI

AI layer

Inference and lightweight fine-tuning run directly on participating devices. Models, datasets and outputs stay auditable end-to-end.

Verification

Proof & reputation

N-way task redundancy with a majority-hash quorum. Play Integrity and node reputation keep the network honest without a gatekeeper.

40–70%
Cost savings vs AWS / Google
99.999%
Availability guarantee
60%
Lower CO₂ emissions
Provider settlement

BSAI — internal coordination unit.

BSAI is an internal accounting unit used inside the VylexAI coordinator to attribute compute work to participating devices. Architecturally comparable to Helium Data Credits or Render Credits — operated centrally by Dudaev Systems UG.

Role in the network

  • Coordination — internal bookkeeping for distributed compute jobs
  • Settlement — providers paid in EUR by Dudaev Systems UG based on BSAI accrued
  • Transparency — every credit verifiable in the coordinator ledger
  • Centrally operated by Dudaev Systems UG (Cottbus, Germany)

Operational properties

  • ✓Internal accounting unit — used by Dudaev Systems UG only
  • ✓Provider-bound — earned per device, settled per Dudaev contract
  • ✓EUR-denominated payouts via SEPA
  • ✓Centralised PostgreSQL ledger · audit-ready
Centrally operated by Dudaev Systems UG, Cottbus, Germany. Public compliance brief available on request.
Roadmap

Three phases. One network.

Launch the mobile layer first. Prove the economy. Scale outward.

  1. Phase 1 — nowLive01 / 03

    Mobile Network Launch

    Smartphones and tablets join the network. Devices contribute AI compute and encrypted storage. Users earn BSAI for every verified contribution.

  2. Phase 202 / 03

    Desktop Expansion

    PCs and laptops plug in as higher-performance nodes — bigger models, heavier AI workloads, more storage capacity.

  3. Phase 303 / 03

    Enterprise Layer

    APIs for companies. An AI training marketplace. Storage-as-a-service. Enterprise dashboards. The network monetized end-to-end.

Team

Built in Germany.

Engineering across Cottbus, Berlin, and Frankfurt. Inside the EU AI Act framework. The VylexAI project is supported by the European Union (ESF+ 2021–2027), the Ministry of Economic Affairs, Labour and Energy of Brandenburg (MWAE), and the Investitionsbank des Landes Brandenburg (ILB), within the framework of StartupLausitz and BTU Brandenburg University of Technology.

Adlan Dudaev
Adlan Dudaev
Founder · Chief Technology Officer
📍 Cottbus, Germany
Ananya Sai Tippani
Ananya Sai Tippani
AI & Machine Learning Engineer
📍 Cottbus, Germany
Hammad Hassan Bajwa
Hammad Hassan Bajwa
Software Engineer (AI / Backend)
📍 Frankfurt, Germany
Riyabrata Mondal
Riyabrata Mondal
Lead Full-Stack & Platform Engineer
📍 Cottbus, Germany
Vishva Hirenkumar Jani
Vishva Hirenkumar Jani
AI Infrastructure & Backend Engineer
📍 Cottbus, Germany
Supported by
Startup Lausitz
Consulting partner
Brandenburgische Technische Universität
Cottbus–Senftenberg

Join the network.

Android beta is shipping. Read the code, follow the progress, or put your phone to work.

GitHub →Join DiscordTalk to the team