OORT1: Privacy First, Cost Conscious

1
OORT was formerly called “Computecoin Network.”. The current version is published on Dec. 29th, 2023.
 1

C ONTENTS

I Executive Summary 2

II Introduction 4
II-A Technical Challenges . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
II-B OORT overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5

III OORT System Architecture 7

III-A Decentralized infrastructure layer . . . . . . . . . . . . . . . . . . . . . . . 7
III-A1 PoH consensus protocol of Olympus protocol . . . . . . . . . . 7
III-A2 Interoperability of Olympus protocol . . . . . . . . . . . . . . . 9
III-A3 Advanced coding technology . . . . . . . . . . . . . . . . . . . 10
III-B AI builder service layer . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11

IV Competitive Landscape 11

V Business Model 12
V-A Who are the clients? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
V-B The Revenue Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12

VI The Tokenomics 13

VII Partners and Customers 14

VIII Publications 15
 2

I. E XECUTIVE S UMMARY

1) Who Is OORT:
OORT is a decentralized cloud designed for data privacy and cost savings. By integrating
global compute and storage resources, OORT provides AI solutions to enhance business
operations

2) Opportunities:
The demand for affordable computing power is skyrocketing in emerging fields such as
AI-generated content (AIGC), Web3, and the Metaverse. Computing power is becoming
the “oil” of future human society. Meanwhile, with growing concerns about data privacy,
applications will increasingly focus on ensuring users’ privacy in data storage and pro-
cessing. Thus motivated, a decentralized data infrastructure with enterprise performance is
needed to address the above two demands.

3) Solution Summary:
The OORT solution comprises two essential layers: the decentralized infrastructure layer
and the generative AI builder platform. The infrastructure layer encompasses a global
infrastructure with an innovative three-tier node architecture. This architecture enables
performance comparable to AWS, supporting exceptional data storage, integration, pro-
cessing, and analytics services. Operating over a decentralized network, this layer ensures
privacy and unparalleled cost savings. Meanwhile, running on the OORT intfrastructure
layer, the generative AI builder platform allows businesses to incorporate factually accurate,
versatile, and secure AI into their operations in minutes. The OORT solution empowers
various Web3 applications such as Game-Fi, Social-Fi, DeFi, NFT, and Web2 applications
to build and scale their operations effortlessly.

4) Products:
a) OORT Storage: Decentralized and enterprise-grade object storage with unparalleled
privacy and up to 80% cost savings.
b) OORT Compute: Decentralized compute engine that empowers machine learning and
all types of data processing functionalities with remarkable cost efficiency.
 3

c) OORT AI: a generative AI agent builder platform, allowing businesses to seamlessly

incorporate factually accurate, versatile, and secure AI into their operations in mere
minutes, not months. No coding or AI specialists are needed.

5) Project highlights:
a) Technical advantage: Continuous research and development since 2018. Core tech-
nologies backed by academic publications and two U.S. patents.
b) Multi-million revenue since November 2021.
c) Strong execution team with core members from world-renowned organizations, such
as Columbia University, Qualcomm, AT&T, J.P. Morgan, etc.
d) World-recognized clients/partnerships: DELL, Lenovo, Tencent Cloud, Alibaba Cloud,
Binance, Cardano, etc.
e) Top-notch advisory board: Dave Mollenhoff (co-founder & ex-CTO of salesforce.com),
Kevin Henshaw (COO & President of Together Labs), Mitch Kennedy (Dell Global
Loyalty Strategy Lead), Steve Liu (Professor at McGill Unviersity, Fellow of the
Canadian Academy of Engineering.), etc.
 4

II. I NTRODUCTION

The emergence of Web3, Metaverse, AIGC and more cutting-edge technologies marks a
historic turning point. Users become more eager for:
1) Low-cost, massive & accessible computing power
The demand for affordable computing power is skyrocketing in emerging fields like Ar-
tificial Intelligence Generated Content (AIGC), Web3, and Metaverse. Computing power
is the “oil” of future human society. Projections for global cloud services market growth
are $596B by 2027 with 14.6% CAGR. AI/ML market size over the next five years is
projected at $200+ by 2027.
2) Data privacy & security
With the growing concerns about data privacy and regulations, applications will focus on
ensuring privacy in data processing and storage. The concentration of sensitive data makes
it an attractive target for cyberattacks and data breaches. A single point of failure creates
vulnerabilities, potentially leading to widespread service disruptions and data loss.
However, cloud service providers like AWS, Azure, and Google Cloud have many limitations in
costs, data privacy, security, sovereignty, and transparency due to the “Nature of Centralization.”
OORT is leading the change. OORT’s decentralized data infrastructure ensures absolute data
privacy and significant savings. In addition, OORT provides an AI-powered data interaction
experience, redefining the data cloud for the next generation.

A. Technical Challenges

Before realizing our vision of revolutionizing the data infrastructure, we need to overcome a
handful of challenging but conquerable obstacles, as discussed below.
1) Not user-friendly
Usability is a critical prerequisite to mainstream adoption. But the fact of the matter is that
Web3 is simply not user-friendly. The vast majority of the world’s population would find
onboarding alone difficult, annoying, and off-putting – to say nothing of actually using
the decentralized web. Web3’s complicated user experience has prevented people outside
of the crypto community from seeing how Web3 can actually generate revenue, benefit
the world, and simply make life easier. We need a set of tools that simplifies the current
workflows in the decentralized world and facilitates new workflows, so that Web3 is no
longer limited to trading. Web3 should be able to cover all of the needs of all of its users.
 5

2) Enterprise-grade performance
Enterprise-grade performance is vital to a useful decentralized data cloud. To compete with
traditional data cloud services, Web3 data cloud must have Web2-comparable capacity
and performance (e.g., scalability, latency, durability, and availability) to handle enormous
volumes of data with guaranteed privacy. For instance, video-streaming and other data-
intensive applications hosted on decentralized networks will need to store and retrieve cold
as well as hot data (respectively, data that’s seldom accessed and frequently accessed data
stored in a database) to facilitate frictionless user experiences. Currently, decentralized
networks can’t host hot data at the internet scale. This is a serious problem. Clients and
users across industries and domains need on-demand, reliable access to data in real-time.
No support for hot data means no video streaming or other kinds of content delivery where
speed is everything.

B. OORT overview

Begun as a university research project in 2018, OORT is founded on the conviction that Web2
users will start migrating to Web3 en masse once the above problems are solved.
OORT is a decentralized data cloud platform that maximizes privacy and cost savings by
integrating global computing and storage resources. It provides enterprise-grade decentralized
cloud-based solutions for generative AI and data-driven businesses. OORT is realizing this goal
via its two layers: the decentralized infrastructure layer and the AI builder service layer. The
details of the two layers will be presented in the following sections.

The three OORT products are:

1) OORT Storage: Decentralized and enterprise-grade object storage with unparalleled privacy
and up to 80% cost savings.
2) OORT Compute: Decentralized compute engine that empowers machine learning and all
types of data processing functionalities with remarkable cost efficiency.
3) OORT AI: a generative AI agent builder platform, allowing businesses to seamlessly in-
corporate factually accurate, versatile, and secure AI into their operations in mere minutes,
not months. No coding or AI specialists are needed.

Some sample business solutions that are built upon OORT products are:
 6

1) Decentralized web hosting: make websites resilient to power outages, server crashes, data
breaches, and censorship.
2) Intelligent personal cloud drive: Securely store, share and manage files with guaranteed
privacy.
3) AI assistant: LLM-powered chatbot tailored to individual and enterprise data, privacy
guaranteed.
4) Blockchain data warehouse: unlock the unlimited potential of Web3 data with unified
on-chain and off-chain data.
5) Data Availability Layer for Blockchains: store stale blockchain data, layer-2 rollup data,
and backup block data.
6) Data market: NFT-based data markets on blockchains.

Fig. 1: OORT two-layer architecture: infrastructure layer and AI builder service layer (aka. Interface layer)
 7

III. OORT S YSTEM A RCHITECTURE

OORT’s system architecture is shown in Fig. 1. It consists of two layers: decentralized

infrastructure layer and AI builder service layer. In this section, we present some technical
details about this two-layer architecture. A more technical system architecture is shown in Fig.
2.

A. Decentralized infrastructure layer

The objective of the infrastructure layer is to build a global decentralized infrastructure with
AWS-comparable performance (otherwise, no possibility of massive adoption). This layer is
the foundation for the data cloud. To achieve AWS-comparable performance, we designed an
innovative three-tier node architecture: super nodes, edge nodes, and backup nodes.
• super node: public and private cloud servers which have superior storage and compute
power.
• edge node: nodes deployed close to the user that forms a peer-to-peer decentralized edge
network (DEN), which is crucial to ensure the scalability of the network.
• backup node: cold-data storage nodes integrated from the existing storage network such as
filecoin, storj, crust and arweave.
The OORT Layer-1 Olympus Protocol, operating on the infrastructure layer, seamlessly in-
tegrates the three-tier nodes worldwide. Olympus Protocol is an original DAG-based layer-1
blockchain that provides consensus and incentives to geo-distributed infrastructure nodes around
the world. One consensus algorithm of Olympus Protocol, the patented Proof of Honesty (PoH),
guarantees the authenticity and honesty of these geo-distributed infrastructure nodes, enabling
businesses to run applications on a trustless decentralized infrastructure.
In what follows, we briefly discuss the key technologies that empower OORT’s superior
performance and the growth of OORT ecosystem: PoH consensus protocol of Olympus protocol,
interoperability of Olympus protocol and advanced coding technologies.
1) PoH consensus protocol of Olympus protocol: The key to implementing a highly scalable
decentralized system is a well-designed resource authentication and management system. Re-
sources typically refers to any physical elements (e.g., bandwidth, storage space) that execute
tasks. At first, the resource contributed by the nodes has to be authentic, which is the foundation
for any decentralized network. For instance, any virtualization technology that forfeit the capacity
 8

Fig. 2: System architecture. Reference: [1] Open Source: https://github.com/OORT-tech [2] Yellow paper: Olympus
Protocol: Consensus and Security [3] Patent: “Methods and Apparatus for Verifying Processing Results and/or Taking
Corrective Actions in Response to A Detected Invalid Result”, S.N. 16/370,629, April. 6th, 2021 [4] Paper: “Cost-
Efficient Hardware Spoofing Detection in Decentralized Shared Computing Network”, Submitted to Sigmetrics.
[5] Paper: “EntrapNet: a Blockchain-Based Verification Protocol for Trustless Computing”, IEEE Journal Internet
of Things, 2021. [6] Paper: “Distributed LDPC Coding Scheme for Low Storage Blockchain Systems”, IEEE
Journal Internet of Things 7 (8), 7054-7071, 2020. [7] Patent: “Methods and Apparatus for Performing Distributed
Computing using Blockchain”, US16/274,178, Aug. 31st, 2021.

of resources should be disallowed. Secondly, for the sake of efficiency and fairness, resources in
a highly scalable system should always be supplied to users with higher utility. However, these
are difficult problems in a decentralized environment due to the lack of global information. From
a game-theoretic point of view, the nodes will schedule resource based on their local knowledge
 9

and with their best personal interest. Therefore, the outcome can not be socially optimal, thus
limiting the network scalability.
To deal with this problem, the OORT team invented PoH consensus protocol, a collection of
decentralized algorithms, for resource authentication and allocation. In particular, as one of the
first practical solution to the long-standing computer science problem “verifiable computing”, the
PoH consensus protocol first verifies that no nodes are cheating on the resource they contribute.
Next, the PoH consensus incentivizes the selfish nodes to act as to optimize the social goals. For
example, in the case that a node receives a file downloading task, a block (a group of transactions)
in the PoH consensus protocol is generated once all the pieces (one transaction for each file
piece transfer) of the file are received by the node. The reward will be proportionally shared
by the nodes who contributed the pieces. With PoH consensus, the nodes are incentivized to
cache frequently accessed files in proximity and to be deployed in a zone with high and reliable
bandwidth. If all nodes were independently working towards this goal, the network topology
and the scheduled contents will be optimized in a decentralized manner.

Fig. 3: Chain interoperability in Olympus

2) Interoperability of Olympus protocol: Olympus protocol is designed to support chain

interoperablity, which enables OORT to integrate Layer 1 Web3 storage protocols such as Filecoin
and Arweave easier, and partner with most of the Layer 1 public blockchains for freely data
 10

transfer in applications.
Chain interoperability in Olympus protocol is achieved by its unique data structure of the
ledger, as shown in Fig. 3. Olympus protocol is split into two layers - the Consensus layer
and the Parachain layer. The consensus layer runs the top consensus of Olympus, which is a
low-complexity and low-latency type of BFT consensus. The reader should read the Olympus
yellowpaper to learn more about the consensus layer. For now, the parachain layer contains two
Olympus’ native chains, one is an EVM-compatible chain to allow developers migrating and
deploying Ethereum smart contracts in OORT with zero learning curve, and the other is a chain
implementing the PoH consensus protocol. Blockchains with different consensus algorithm, such
as Binance Smart Chain, Solana, Filecoin, Arweave etc., can be added to the parachain layer.
The messages in the consensus layer are called blocks. A block connect to previous blocks
and forms a DAG. The Olympus consensus determines a linear order of the blocks, though
they are tangled. The order of the blocks are marked on Fig. 3 as an example. Each block in
parachains is linked from a block in Olympus, and then the order of the transactions among the
parachains are subsequently determined, which is the fundamental step of implementing chain
interoperability.
To understand how the chain interoperability can benifit OORT, on one hand, the cross-chain
feature allows OORT to adopt any popular payment methods in the crypto space. OORT users
don’t have to be aware of the existence of different tokens in different Layer 1 protocols. The
tokens will be automatically purchased and consumed by the cross-chain protocol in Olympus.
On the other hand, the cross-chain feature enables the developers to build applications on Web3
data in a much broader scope. Data as an important asset in Web3 can be easily transferred
or exchanged across different blockchains in OORT. Many novel applications, such as Web3
data marketplace and data warehouse, will be built by developers as they explore the infinite
opportunities brought by OORT.
3) Advanced coding technology: When a user wants to upload a file to OORT, the file is
split into small fixed-size pieces known as “shards”. The total number of “shard” is determined
by the size of the file. Each shard is encrypted using client-side encryption AES-256, ensuring
that only the file owner can access the content. After encryption, the file is processed using
a technique called erasure coding such as polar code and LDPC code. This process generates
additional “parity” shards, which provide redundancy and allow the file to be reconstructed even
if some shards are lost or unavailable. These shards (both the original and the parity ones) are
 11

then distributed across OORT decentralized infrastructure nodes. The shard IP-based distribution
is recorded in the decentralized database system in OORT. When the user wants to retrieve their
file, the client-side application requests the shards from the corresponding nodes by querying
the database. As the result of the erasure coding technology, the file can be reassembled and
decrypted even if some shards are not available due to node failure or network congestion. This
significantly improves the reliability and availability in a unstable decentralized network.

B. AI builder service layer

OORT AI revolutionizes the way individuals interact with their data, particularly targeting non-
technical users. Similar to the GPT model, OORT AI provides a user-friendly conversational AI
builder platform that simplifies data management and interaction. By integrating OORT AI into
the business, the user experience of using generative AI is redefined to a straightforward process:
“upload your data, converse with your data.”
OORT AI is constructed based on its own fine-tuned large language model, derived from the
open-source Large Language Model (LLM). Extensive research and implementation efforts have
been dedicated to optimizing OORT AI, ensuring both cost-effectiveness and high-performance
scalability. The scalability of OORT AI addresses potential concerns related to cost and perfor-
mance.
The three key features of OORT AI are as follows:
1) Customization: OORT AI takes customization to new heights, using your data to cater to
your unique needs. This breakthrough approach opens up fresh possibilities to extract the
most value from your data, aligning with our vision of a decentralized data economy.
2) Data Sovereignty: With OORT’s unique decentralized infrastructure, we prioritize data
ownership. Here, your data isn’t just securely stored, it’s rightfully yours, driving our
mission to revolutionize human-to-data interaction.
3) Community at its core: Experience the power of community. Empower your business by
drawing on high-quality data from your users to elevate OORT AI capabilities. Incentivize
community contributions and build a secure, customized chatbot for your business.

IV. C OMPETITIVE L ANDSCAPE

So far, no company has positioned itself as a decentralized data cloud while offering the same
services as OORT. OORT’s offerings overlap with both Web2 and Web3 cloud infrastructure
 12

companies. Web2 companies include AWS, Google Cloud, Azure, and others. Web3 companies
encompass Filecoin, Dfinity, and more. However, OORT distinguishes itself as an AI-focused
decentralized cloud. Its services are highly concentrated on cost savings and privacy. OORT
provides enterprise-grade decentralized cloud-based solutions specifically tailored for generative
AI and data-driven businesses. Below is a table illustrating the competitive landscape

V. B USINESS M ODEL

A. Who are the clients?

OORT offers comprehensive data cloud solutions that cater to a wide range of clients, both
in the Web2 and Web3 domains, regardless of their size. As the Web3 community is still in its
early stages, we have identified two primary client segments that can swiftly generate revenue
for us.
1) Web3-native projects: This category encompasses projects inherently built for the Web3
ecosystem, requiring decentralized data solutions for their web3 products, including project
websites and databases. Our experience in acquiring clients from this segment has revealed
that the client acquisition cost is exceptionally low, and the onboarding process is rapid
and efficient.
2) Web2 enterprises transitioning to Web3: These companies are highly motivated to expand
their business operations into the Web3 realm. The educational cost associated with these
clients is relatively lower compared to those with little knowledge of Web3. Our focus is
primarily on small to medium enterprises (SMEs) that often face challenges in receiving
adequate support from traditional cloud service providers like AWS. SMEs, being cost-
sensitive, are unable to benefit from the same level of discounts offered to larger enterprises
by AWS. Leveraging our founding team’s established track record and reputation, we are
in a favorable position to secure these clients.

B. The Revenue Model

1) Service Fee: Our primary business is to provide a comprehensive data cloud and one-stop
generative AI solutions to meet the specific needs of the clients. We will offer pay-as-
you-go and subscription pricing in the suite of OORT products to our clients. The price
is on average 50% − 80% off compared to that of AWS. The big savings come from
 13

leveraging under-utilized resources in the infrastructure layer and the optimized utilization
of the three-node architecture for any application.
2) Licensing Fee: The providers, nodes in our decentralized network, are charged an annual
licensing fee for their connection to OORT. Note that the benefits they earn from providing
their resources to our network will be far greater than the licensing fee.

Fig. 4: Token Allocation.

VI. T HE T OKENOMICS

OORT utilizes the utility token native to the OORT ecosystem. The utility token is used for
service payment, staking, governance. The token allocation is illustrated in Fig. 4.
1) Service Payment: Users can always pay fiat currency (via credit card) for different services
in the OORT ecosystem. However, a discount is usually given to the users who use OORT
utility tokens for service payments.
2) Staking: OORT offers two ways for token holders to stake tokens:
a) Stake tokens with staking pools. One can delegate tokens to node providers via OORT
bonding pools. Node providers will use staked tokens as mining collateral. As a result,
you will receive a share of mining rewards from the provider.
b) Become a node provider. Stake tokens as mining collateral.
 14

3) Govenance: Governance is one of the primary functions of the OORT ecosystem. Users
submit proposals and vote on topics like network parameters, slashing mechanisms, and
service fees. A proposal must be submitted with a certain amount of deposit, which is
fully refundable once the proposal passes. If the proposal fails to pass, deposits will be
distributed to the mining rewards pool for node providers.
The OORT team will use a certain amount of net revenue to regularly buy back and burn
its utility tokens to reduce the total token supply on a quarterly basis and thus gradually and
continuously increase token value. The token burning cap is 40% of the total token supply.

Fig. 5: Selected partners and collaborators.

VII. PARTNERS AND C USTOMERS

Our team’s track record and reputation have helped us build a very strong ecosystem of
partnerships and secure world-reputable customers. These partners and customers include top
universities in the U.S., legacy Internet giants, innovative Web3 projects, data centers, and media
platforms around the world. Our partnerships and early customers have significantly enhanced
our branding, user acquisition, fundraising, and system development. Below are some samples
of our partners and customers:
1) Lenovo (Customer)
 15

Lenovo Image has integrated OORT TDS solutions into its products to enhance its customer
service. OORT solution will be extended to Lenovo’s core PC business very soon. Annual
PC sales are 60 million units.
2) Dell Technology (Partner/Distributor)
DELL & OORT loyalty program launched on July 7th, 2023. This program will promote
OORT products to 10+ million small and medium companies.
3) Tencent Cloud (Partner/Distributor)
Tecent Cloud data center has joined OORT infrastructure network as OORT’s super nodes.
OORT products have been promoted to the Tencent Cloud ecosystem.
4) Alibaba Cloud (Partner/Distributor)
Integrate OORT products into AliCloud marketplace for their 4+ million customers in
Asia.
5) Binance BNB Chain (Partner/Distributor/Customer)
Binance BNB Greenfield (a decentralized storage protocol) has integrated OORT decen-
tralized cloud solution, as a service provider, to empower BNB ecosystem projects. In
addition, OORT has been a cloud service provider for BNB ecosystem.
Fig. 5 shows the ecosystem of OORT partnership.

VIII. P UBLICATIONS

Academic Papers:
1. C. Li, L. Zhang and S. Fang “EntrapNet: a Blockchain-Based Verification Protocol for
Trustless Computing”, IEEE Journal Internet of Things, 2021.
2. H. Wu, A. Ashikhmin, X. Wang, C. Li, S. Yang and L. Zhang, “Distributed LDPC Coding
Scheme for Low Storage Blockchain Systems”, IEEE Journal Internet of Things 7 (8), 7054-
7071, 2020.
3. “Cost-Efficient Hardware Spoofing Detection in Decentralized Shared Computing Network”,
Submitted to Sigmetrics.
Patents:
1. C. Li, L. Zhang and S. Yang, “Methods and Apparatus for Performing Distributed Com-
puting using Blockchain”, US16/274,178, Aug. 31st, 2021 (Granted).
 16

2. C. Li, L. Zhang and S. Yang. “Methods and Apparatus for Verifying Processing Results
and/or Taking Corrective Actions in Response to A Detected Invalid Result”, S.N. 16/370,629,
April. 6th, 2021 (Granted).
 17

Built by innovators, for innovators.