Sahara AI is developing a purpose-built Layer 1 blockchain that aims to bring the entire lifecycle of artificial-intelligence assets - data, models and autonomous agents - directly on-chain. By combining an EVM-compatible proof-of-stake chain with encrypted vaults, an integrated compute hub and a licensing marketplace, the project hopes to give contributors transparent ownership, verifiable provenance and a fair share of downstream revenue. A $43m Series A round last August provided the runway, and the public SIWA testnet keeps the network on course for a mainnet launch in 2025.
Pier Two Testnet Participation
Pier Two recently participated in the Sahara AI testnet and has written open-source code to assist with the deployment and maintenance of validator software in the Sahara AI ecosystem. Please reach out to Pier Two to learn more!
Deployment Code
- name: "Sahara AI Node Deployment"
hosts: all
become: yes
vars:
datadir: "/data1/sahara"
moniker: "{{ app.vars.moniker }}"
external_ip: "{{ app.vars.external_ip }}"
tasks:
- name: Check if Sahara Testnet node containers are running
command: docker compose ps --services --filter "status=running"
args:
chdir: "{{ datadir }}/setup-testnet-node"
register: compose_running
changed_when: false
failed_when: false
- name: Stop existing Sahara Testnet node containers
command: docker compose down
args:
chdir: "{{ datadir }}/setup-testnet-node"
when: compose_running.stdout != ""
- name: Start Sahara Testnet node with Docker Compose
command: docker compose up -d
args:
chdir: "{{ datadir }}/setup-testnet-node"
What Sahara AI is trying to solve
Today’s large-model economy is dominated by a few cloud vendors that control training data, compute and distribution. Sahara’s thesis is that value and attribution should flow back to the people who supply those resources. The network is being designed to
- record clear provenance for every dataset, model snapshot and agent
- protect privacy through secret-sharing and differential-privacy techniques in encrypted vaults
- route usage fees automatically to everyone in an asset’s dependency graph
How the stack fits together
Sahara Blockchain
The settlement and registry layer uses fast-finality consensus and AI-specific pre-compiles to verify proofs of training or inference without flooding the chain with raw data.
Vaults
Encrypted storage keeps proprietary data and model weights confidential until licences are granted. Access can be scoped to a specific time window or workload.
Compute Hub
A unified interface lets developers rent GPU capacity from cloud providers, specialist marketplaces and community nodes, all without leaving the ecosystem.
Marketplace
Datasets, models and turnkey agents can be listed here with royalty splits encoded in smart contracts, so contributors are paid whenever their work is used.
Recent milestones and what’s next
- August 2024 – Series A closes, led by Pantera Capital, Binance Labs and Polychain
- December 2024 – Data Services Platform opens for community data collection and validation
- March 2025 – SIWA public testnet goes live for developers to register and license AI assets
- Second half 2025 – Targeted mainnet launch with the first cohort of production validators
Pier Two’s role on the testnet
Pier Two is currently operating validator and node infrastructure on the SIWA testnet, applying the same institutional-grade practices we use across other networks. Our contribution includes hardened cloud architecture, SOC 2-aligned operational controls and real-time performance monitoring. We’re excited to join the Sahara AI community and provide the reliability needed to gather accurate performance data ahead of mainnet.
Why Sahara AI is worth keeping track of
Sahara AI sits at the intersection of blockchain governance and applied AI. If its on-chain provenance and automated royalty model work as intended, the network could become an asset registry and settlement layer for builders who need clear licensing and verifiable audit trails. For node operators and stakers, it also represents a chance to diversify rewards beyond standard transaction fees by participating in revenue streams tied to data and model usage.
Looking ahead
The public testnet will continue to harden core protocols through 2025 while token economics and validator parameters are finalised. Mainnet launch is pencilled in for the second half of the year, followed by releases of an SDK for autonomous agents and expanded compute-for-stake incentives. Pier Two plans to remain active in testing and, when mainnet goes live, to extend our professional-grade services to the production network.