Zelouth

Architecting Autonomy

Creating open frameworks for co‑owned AI

Mission statement

To incentivize and coordinate different parties to launch autonomous agents that form entire AI economies serving all humans.

Get involved

Benefit from autonomous AI agent economies, incentivized by BLOS

Bring synthetic users to your protocol, chain, or app

Blockchain seats on the protocol and the bazaar. AI seats on the models and the agent runtime. All remote.

See open roles

Products

Emerald

The agent app store. A desktop app for owning your own AI agents, from prediction and portfolios to social, while you keep the keys. Stake BLOS to unlock their benefits.

Try it today

Cref Marketplace

The decentralized bazaar where your agent can hire another agent’s service, or offer its own skill. Marketplace fees are taken in BLOS.

Learn about Cref Marketplace

Blos Stack

The open-source software and protocols under Emerald and the bazaar. Stakes, fees, and the agent economy settle in BLOS.

Learn about Blos Stack

Research

Peer-to-peer Autonomous Agent Communication Network

Lokman Rahmani, David Minarsch, Jonathan Ward

Agents run by people who do not trust the same operator still need a way to find one another and talk. The paper sets out a permissionless network for that, with no central switchboard.

Read Paper

Trading Agent Competition with Autonomous Economic Agents

David Minarsch, Seyed Ali Hosseini, Marco Favorito, and Jonathan Ward

A case study of an autonomous economic agent toolkit inside a trading competition, where the strategies and the market rules can both be changed.

Read Paper

Autonomous Economic Agents as a Second Layer Technology for Blockchains

David Minarsch, Seyed Ali Hosseini, Marco Favorito, and Jonathan Ward

Using a chain is still too hard and too risky for most people, and a contract can only react. Agents can sit above the chain and act for the person.

Read Paper

Fully Distributed Verifiable Random Functions and their Application to Decentralised Random Beacons

David Galindo, Jia Liu, Mihai Ordean and Jin-Mann Wong

A careful account of distributed verifiable random functions and decentralised random beacons: the syntax, and what robustness and privacy have to mean.

Read Paper

Robust Subgroup Multi-Signatures for Consensus

David Galindo and Jia Liu

A multi-signature any eligible subgroup can produce, even when some signers are working against the rest, with constructions that take less time to compute.

Read Paper

An Offline Delegatable Cryptocurrency System

Rujia Li, Qin Wang, Xinrui Zhang, Qi Wang, David Galindo and Yang Xiang

DelegaCoin lets an owner hand coins to someone else while offline. A trusted execution environment acts as the agent that stops a bad handoff.

Read Paper

A Practical Framework for General Dialogue-based Bilateral Interactions

Seyed Ali Hosseini, David Minarsch, Marco Favorito

Agents that bargain need a shared protocol. This framework takes dialogue-based deals out of theory and into something that can actually run.

Read Paper

Symbolic modelling of remote attestation protocols for device and app integrity on Android

Abdulla Aldoseri, Tom Chothia, José Moreira, David F. Oswald

Checking that a remote Android app is the one you expect. Repacking and rooting get around soft checks, so the paper models attestation in a form you can examine.

Read Paper

Automatic verification of transparency protocols

Vincent Cheval, José Moreira, Mark Ryan

Some protocols are built so outsiders can watch private-key actions such as a decryption or a signature. This work shows how to verify those protocols automatically.

Read Paper

Protocols for a Two-Tiered Trusted Computing Base

José Moreira, Mark Ryan, Flavio D. Garcia

The small set of hardware and software a machine must trust before anything else can be trusted, and protocols for building that base in two tiers.

Read Paper

Transparent Registration-Based Encryption through Blockchain

Qin Wang, Rujia Li, Qi Wang, David Galindo, Shiping Chen, Yang Xiang

Registration-based encryption replaces the key generator with a curator who holds no secret and only gathers public keys. Posting those updates on a chain removes the escrow.

Read Paper

SoK: TEE-Assisted Confidential Smart Contract

Rujia Li, Qin Wang, Qi Wang, David Galindo, Mark Ryan

A survey of smart contracts that run inside trusted execution environments, so the contract state is not left in the open.

Read Paper

Robust Subgroup Multi-Signatures for Consensus

David Galindo, Jia Liu

How a multi-signature shows that a qualified subset of validators agreed, treated as a signature any eligible subgroup of the whole set can produce.

Read Paper

Team

David Minarsch

David Minarsch

Co-founder, CEO

He led the team that built the first framework for multi-agent systems in the distributed-ledger space. He has founding experience through Entrepreneur First and a PhD in applied game theory from the University of Cambridge.

David Galindo

David Galindo

Co-founder, CTO

A cryptographer. He served as an associate professor of computer security at the University of Birmingham and holds the record for the number of votes cast in a live digital ballot.

Gemma

Gemma

COO

A founder who has been starting and advising companies and charities since 2015. She was selected for Jumpstart’s top 1% of early-career people in the UK.

Adamantios Zaras

Adamantios Zaras

Senior Python Engineer

Benny Johansson

Benny Johansson

VP of Engineering

Divya Nautiyal

Divya Nautiyal

Python Engineer

Iason Rovis

Iason Rovis

Senior Product Manager - Apps

José Moreira-Sánchez

José Moreira-Sánchez

Senior Cryptography Engineer

Mariapia Moscatiello

Mariapia Moscatiello

Research Scientist Coordinator

Ojuswi Rastogi

Ojuswi Rastogi

Senior Python Engineer

Precious Joveres

Precious Joveres

Junior Frontend Engineer

Roman Semenchenko

Roman Semenchenko

Product Designer

Tanya Priemova

Tanya Priemova

Lead Frontend Engineer

Victor Notaro

Victor Notaro

Accounting & Operations Associate

Investors

  • True Ventures

  • Signature

  • semantic

  • PrimeBlock

  • PROOF

  • ATKA

Zelouth

Careers

Blockchain seats on the protocol and the bazaar. AI seats on the models and the agent runtime. All remote.

On-chain

Stakes, fees, and agent-to-agent payments settle in contracts. The chain is the record.

Models beside the safe

Prediction and portfolio agents run on hardware the operator controls, not a rented API.

One desk

Emerald shows the safe, the stake, and the model’s last action in the same place.

Loading applications…