A LGORITHMIC T RADING P LATFORM

AND

X EN T OKEN

W HITEPAPER (D RAFT )

Shuvro Biswas
MS A PPLIED M ATHEMATICS
(C O -F OUNDER , C HIEF E XECUTIVE O FFICER & C HIEF A RTIFICIAL I NTELLIGENCE O FFICER )

Robin Raymond
P H D M ATHEMATICS
(C O -F OUNDER & C HIEF T ECHNOLOGY O FFICER )

Kirill Kanshin
P H D P HYSICS
(C O -F OUNDER & C HIEF D ATA S CIENTIST )
© 2018 C RYZEN LLC
ALL RIGHTS RESERVED

WWW. CRYZEN . COM

Latest Changes: July 31, 2018

 C ONTENTS

1 Introduction and Problem Statement . . . . . . . . . . . . . . . . . . . . . . 5

2 Cryzen Trading Platform . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

2.1 Data Collection and Management 7

2.2 Backtesting and Live Deployment 8

2.3 Code Studio 10

2.4 Trading Parameter Optimization 11

2.5 Machine Learning 12

2.6 Community Algorithms 13

2.7 Design Considerations 13

2.7.1 Programming Language: Python . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13

2.7.2 Backtesting System: Event Driven + Full Orderbook Level Simulations . . . 14

4

3 XEN Token . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15

4 XEN Token Revenue Sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17

4.1 VaaS Fees 17

4.2 Premium Service Fees 18

4.3 Zen Fund Trading 19

5 Token Sale . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20

6 Cryzen DAPP Voting System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23

7 Future Vision . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24

8 Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
 I NTRODUCTION AND P ROBLEM
S TATEMENT

The cryptocurrency exchange and price space is highly inefficient and volatile. In
a single day, prices may swing upwards of 10% on market news and sentiment.
In order to take advantage of interexchange inefficiencies (e.g. arbitrage between
exchanges), and intraexchange inefficiencies (e.g. price volatility, price/informa-
tion lags, etc), a robust algorithmic trading platform is required supporting both
backtesting and live deployment of trading algorithms.

In this whitepaper we describe the state of the art Cryzen trading platform
and accompanying XEN token. The Cryzen trading platform consists of two parts:
backtesting and live deployment. Backtesting consists of running or simulating a
trading algorithm on past price, social media, and other data, while live deployment
consists of running that same trading algorithm (without modification) on live
markets. Our trading platform (described in more detail in section 2) abstracts
away minute detail related to 1) the need for consistent and reliable data, and 2)
the need to interact with various cryptocurrency exchanges.

This simplification layer allows users to develop trading strategies at a very

high level (take data feeds as input, and produce trading actions as output), and
immediately deploy them to the live market upon satisfactory backtesting results.
Specifically, users can develop algorithms at the level of Python code, or tweak/edit
template strategies from a simplified UI requiring no coding knowledge, allowing
both programmers and non-programmers to benefit from our system.

Successful algorithmic trading contributes to an overall more efficient cryp-

6 Chapter 1. Introduction and Problem Statement

tocurrency market (less arbitrage between exchanges, less unnecessary volatility,

and faster/greater reflection of news and social media sentiment in cryptocurrency
prices). Thus while profiting, our users in the aggregate also bring greater stability
and efficiency to the cryptocurrency price space.

As of the publication date of this whitepaper, our trading platform boasts

the highest frequency trading option available (at up to 1s intervals) and orderbook
level data at both the backtesting and live deployment level (relative to other
trading platforms). Further details are provided in the trading platform section.

XEN token (an ERC20 token on the Ethereum blockchain) will be one of the
first dividend tokens to be deployed and connected to a private system for the
purposes of distributing profits to token holders. XEN token will distribute 40%
of profits made from three revenue sources: 1) value as a service fees (a 5% fee on
user profits), 2) premium service fees, and 3) proceeds from trading the Zen fund, a
special fund that will be powered by the XEN token sale and traded by an in-house
(Cryzen) quant team.

All revenue and profit data from the aforementioned sources will be made
available to our users publically. A smart contract / oracle system will then au-
tonomously make use of this public information to distribute 40% of profits to XEN
token holders on a periodic basis (profits are calculated in USD, and the equiva-
lent is paid out in ETH) - the other 60% of profits will be used to fund ongoing
operations of Cryzen LLC.

At a high level, purchasers of XEN tokens are contributing to the develop-

ment and running of trading algorithms which in trying to maximize profit also
help create more efficient and stable cryptocurrency markets.
 C RYZEN T RADING P LATFORM

The Cryzen trading platform is the foundation of all trading algorithms. There
are two major aspects to the Cryzen trading platform: 1) data collection and man-
agement, and 2) backtesting of algorithms on past data, and live deployment of
those same algorithms to the market. The Cryzen trading platform is accessed via a
web interface called the code studio where all algorithms are developed, backtested,
visualized, and deployed. We discuss all of the major aspects of the Cryzen trading
platform and code studio in the sections below.

2.1 Data Collection and Management

Our data collection systems record data from several exchanges, including but not
limited to: Coinbase Pro, Binance, Gemini, and Okex (more will be supported on
an ongoing basis over time). For exchanges that support websocket connections,
we collect orderbook level data at 1 sec intervals. For exchanges that support rest
api calls only, we collect orderbook level data at 5 sec intervals.

While the cryptocurrency markets are inefficient enough that price data
alone is sufficient for profitable trading strategies (e.g. arbitrage, classic momen-
tum strategies, etc), we intend to further support social media, news, event, and
blockchain level data sources to allow for a full spectrum of trading capabilities. To
this end, we will collect the following data:
 8 Chapter 2. Cryzen Trading Platform

I Twitter data with relevant hashtags

I Reddit data from relevant subreddits
I Google trends data for relevant search phrases
I Cryptocurrency event data (new exchange listings, cryptocurrency forks,
network upgrades, and more)
I Blockchain level data (confirmation times, large transfers, number of unique
wallets, and more)

Users will be able to backtest on all of these data sources as well as run
algorithms live on these feeds seamlessly.

2.2 Backtesting and Live Deployment

The backtesting and live deployment system is best described by the diagram below.
Users of the Cryzen trading platform develop (or modify) a trading algorithm A
which takes as input: data feeds + previous feedback, and produces as output:
actions to execute.

Figure 2.1: Trading Platform Flow Chart

All input data feeds are provided to the most recent time point tn . Feedback,
however, always corresponds to either the previous time point tn−1 , or even earlier
time points tn−k , where k is an integer greater than 1.

Immediate or tn−1 feedback relays to algorithm A the immediate conse-

quences of the preceding set of actions it took (e.g. A placed a buy order in the
market at the previous time point, and is now receiving feedback that the order
9 Chapter 2. Cryzen Trading Platform

has been filled). Delayed or tn−k feedback relays to algorithm A the delayed conse-
quences of an earlier set of actions it took (e.g. A initiated a cryptocurrency transfer
from exchange X to Y, k time points ago, and is now receiving feedback that the
transfer has completed).

Actions produced by algorithm A are derived from the following set of

primitives: buy, sell, modify, cancel, and transfer. More complex action types can
be composed from these basic primitives.

Algorithm A can be run on either the backtester or deployed in the live

market (from the algorithm’s point of view there is no difference). For backtesting,
the user will select a time range to backtest over and specify initial balances for the
relevant cryptocurrencies and exchanges involved.

Upon satisfactory back-

testing, the user may deploy A
to the live markets (in fact, for
this to happen, our system re-
quires the user to backtest on
at least 1 month of data first).
Once the user launches A to the
live markets, he or she will be
asked for API keys to the rele-
vant cryptocurrency exchanges
involved. These keys will then
Figure 2.2: System Architecture
be sent to a docker container
where the user’s algorithm A
will run live (isolated and secured from all other users and the rest of our system).
A simple schematic of the system architecture is given in figure 2.2.

Once live deployment of an algorithm is stopped, the corresponding docker

container running it is destroyed and along with it any trace of the user’s API
keys. For technically minded and curious readers, backtests are also run on docker
containers in a similar manner, keeping user developed algorithms secure and
isolated from the rest of our system; this in turn also protects our systems from any
possible malicious harm.
 10 Chapter 2. Cryzen Trading Platform

Figure 2.3: The Code Studio

2.3 Code Studio

Users will develop trading algorithms through a web interface called the code
studio (depicted in figure 2.3). The code studio provides an online / browser
based ide for algorithm development in Python and supports syntax highlighting,
autocomplete, code collapse, autoindent, and more. Furthermore, the code studio
features a mini dashboard which will display trading logs and render/visualize
backtesting results with respect to price and action data (i.e. it will render an
overlay of what a user’s script did on top of the price charts of the cryptocurrencies
involved).

Further, customizable template

algorithms will be available to edit,
backtest, and deploy to live markets.
For convenience and ease of use, we
will provide a “code-free” interface to
these template algorithms called the
Zen interface (depicted in figure 2.4).
The “Zen form” of a trading algorithm
(provided through the Zen interface)
will allow users to edit the algorithm via
input boxes, sliders, dropdown menus, Figure 2.4: The Zen Interface
and more (abstracting away the need to
 11 Chapter 2. Cryzen Trading Platform

view or edit code). In reality, this “code-free” interface to trading algorithms, is just
a high level code editor that manipulates template algorithms behind the scenes
and runs them through the normal machinery.

The Zen form of a trading algorithm can be easily generated using our
parameter tagging library. After development of a trading algorithm, the user can
indicate trading parameters using our parameter tagging syntax. For example,
consider a trading algorithm containing the following variable:

buy_amt = 10

The user/developer of this algorithm can indicate that this is a trading

parameter via parameter tagging syntax as follows:

buy_amt = parameter(10, range=(10, 100)) # tagged parameter

Trading algorithms enhanced with tagged parameters can be systematically

parsed for the parameters it depends on to autogenerate a UI containing edit access
to those parameters.

Cryzen will provide many classic trading algorithms in both code and UI
/ Zen form so that both coders and non-coders can utilize them. This includes
but is not limited to: arbitrage algorithms, momentum algorithms, and portfolio
rebalancing algorithms.

2.4 Trading Parameter Optimization

The Cryzen trading platform provides out of the box trading parameter optimiza-
tion solutions. Once an algorithm’s parameters have been tagged, we have the
ability to parse and understand trading parameters. A simple extension of this capa-
bility is to run many backtests in parallel (using our docker container architecture)
where different trading parameter values are tested in different docker containers
 12 Chapter 2. Cryzen Trading Platform

to find the most optimal configuration of parameter values (most optimal with
respect to “searched” parameter configurations, rather than all possible parameter
configurations). Consider a trading algorithm containing the following lines of
code:

buy_amt = parameter(10, range=(10, 100))

sell_amt = parameter(10, range=(10, 100))

As the trading parameters have been tagged, it is possible for us to conduct a

grid search over the parameter space to find the most optimal (searched) parameter
configuration. The user can specify at a high level, how detailed they want this
search process to be (a more detailed search will use more docker containers, and
the user may be billed higher resource use fees as a result).

2.5 Machine Learning

The Cryzen trading platform will also contain a separate machine learning interface
for the development and training of predictive machine learning models. Using
this interface, users will define some aspect of the future to predict, given some
aspect of the past, and a machine learning model to mediate that prediction.

For example, the user may want to predict the standard deviation of cryp-
tocurrency prices over the next hour (volatility in other words), given the average
price, volume, and twitter sentiment of that cryptocurrency over the last hour,
using a random forest model. The user will first specify the time interval to train
their model over (this includes a start time ts and an end time te for the given
interval). Afterwards, the user will specify the hyperparameters of their model, or
conduct an automated grid search over hyperparameters. Once the user is satisfied
with the performance of a trained model, this trained model can be used on both
the backtester and live markets. To do this, the user would import their trained
model into the Code studio and write trading logic based on the predictions of the
imported model (multiple models can be imported and used in combination in the
Code studio).

Curious readers may wonder how exactly machine learning models are
backtested since training models on a given time interval and backtesting anywhere
 13 Chapter 2. Cryzen Trading Platform

on that same time interval would be biased (i.e. the machine learning model was
trained to perform well on said time interval). To address this issue, we ensure that
backtesting can only be done from te onwards. The Cryzen trading platform will
simply not allow users to backtest imported models on dates prior to te to prevent
biased / unrealistic backtesting results.

2.6 Community Algorithms

We allow users to share algorithms freely to the community. We believe this is cru-
cial to foster innovation, peer review, and thoughtful discourse in the algorithmic
trading community. Interesting / promising community algorithms will be aug-
mented with our parameter markup language and made available to non-coders
through the Zen interface as discussed above.

2.7 Design Considerations

2.7.1 Programming Language: Python

At a high level, the possible programming languages for an algorithmic trading

platform include: domain specific languages (DSLs), scripting languages (such as
Python or Lua), and compiled languages (such as C++, Rust or Haskell). While
DSLs provide the benefit of easy isolation they have the drawback of requiring
users to learn a new language. Furthermore, only features implemented in the DSL
will be available to the user, restricting possible trading algorithm implementations.
Compiled languages have the benefit of potentially resulting in faster algorithms,
but this might be negligible compared to other factors such as colocation. The main
drawback of compiled languages is that the development cycle of an algorithm is
slowed down considerably by lower level syntax and the need for compilation. We
chose Python as the language of choice for the Cryzen trading platform because of
its high level / clean syntax, and the large overlap it has with quantitative finance
and machine learning.

Further, we chose to containerize the execution of backtests and live running

algorithms, to enable us to scale the use of python for mass algorithmic trading.
 14 Chapter 2. Cryzen Trading Platform

2.7.2 Backtesting System: Event Driven + Full Orderbook Level Simulations

The Cryzen trading platform boasts a state of the art backtesting system that takes
into account some of the special challenges present in cryptocurrency trading. As
usual, we enforce the constraint that the same algorithm used for backtesting is
also the same algorithm that is deployed live, and that there is no distinction from
the algorithm’s point of view between the two (this is typical of most backtesting /
live deployment systems).

The Cryzen trading platform is event driven: trading algorithms running on

either the backtester or in the live markets are fed data whenever there is an update
to any relevant data stream (price data, social media data, etc) - these updates are
considered “events”. Our backtesting system feeds “events” to trading algorithms
in the exact same manner as would have occurred in the live markets (but at a
much higher rate). This ensures that there are no systematic differences between an
algorithm running on the backtester vs live from a data / event point of view.

Most backtesting systems work with candlestick data (open, close, high, and
low prices over some time interval - usually 1 minute). These systems make the
assumption that there is unlimited volume on the best bid and best ask (or within a
candle) for purposes of backtesting. Our backtesting system, however, conducts
simulations at the level of orderbooks at up to 1 sec frequencies. This is necessary
for accurate backtesting in the cryptocurrency markets as volumes on the best bid
and best ask in these markets are often quite low and it becomes severely inaccurate
to make an unlimited volume assumption here.

The lower volumes involved in cryptocurrency trading, compared to tra-

ditional financial asset trading, further makes it necessary to simulate the impact
that the trader has on the market (with respect to backtesting). Therefore our
backtesting platform includes “forward propagation” of orders - a way to simulate
the impact of a users trading algorithm on past data and market conditions. For
example, if an algorithm places a large market order to buy 1000 ETH that would
have changed market prices, we take that into consideration and change market
prices accordingly on our backtester. Most backtesters, however, would make the
unlimited volume assumption and execute a 1000 ETH buy order without taking
into account that such an order would not be executable at the best ask and would
end up changing market prices for ETH (making our backtester state of the art for
cryptocurrency backtesting). For purposes of intellectual property, we leave the
exact specifics of forward propagation out of this whitepaper.
 XEN T OKEN

XEN token will be one of the first dividend tokens, distributing profits generated
from Cryzen revenue sources to token holders.

Profits generated by Cryzen will be published weekly to a publicly available

webpage located at cryzen.com/profits. In specific, detailed breakdowns of value
as a service (VaaS) fees, premium service fees, and zen fund trading profits will be
published on cryzen.com/profits (see the XEN Token Revenue Sources section for
details on Cryzen revenue sources).

A vault smart contract / ora-

cle system (connected to our company
ethereum wallet) will check this web-
page on a weekly basis, and using the
reported net profit, send ETH to our dis-
tribution smart contract. Once ETH ar-
rives at the distribution smart contract,
it is locked and cannot be retrieved by
us. At the end of the month, all ETH
locked in the distribution smart con-
Figure 3.1: Smart Contracts
tract is available for withdrawal by to-
ken holders. For top token holders, we
initiate this withdrawal so that the ETH they are owed is sent to their Ethereum
wallet. Snapshots of token balances are taken at the end of each month so that
we have an exact record of how much each user is owed (our snapshots can be
16 Chapter 3. XEN Token

independently verified on the blockchain). A full diagram of all smart contracts

involved in our system is given in figure 3.1.
 XEN T OKEN R EVENUE S OURCES

The three primary revenue streams for Cryzen (40% of which will be distributed to
XEN token holders) include: 1) value as a service (VaaS) fees, 2) premium service
fees, and 3) proceeds from trading the Zen fund.

4.1 VaaS Fees

VaaS fees apply to user profits: if a user has made profit through algorithmic trading
on our platform for a given month (across all trading strategies and algorithms
employed), he or she will be charged 5% on those profits. In specific, user profits
for a given month is calculated as follows:

For the duration of each live algorithm, we record the buy and sell behavior
across all exchanges involved. For each transaction, we record the currency used
for that transaction (amount used), and the currency gained as a result of that
transaction (amount gained). An amount used is always a negative number, and an
amount gained is always a positive number. For all currencies involved, a running
account of the net position is kept (total amount gained + total amount used). At
the end of the month, the net position of each currency is converted into dollars
(based on the current market price of each currency) and summed to yield the user
profit. If the user profit is positive, we charge 5% of it as the VaaS fee for the given
user.
 18 Chapter 4. XEN Token Revenue Sources

4.2 Premium Service Fees

Basic usage of the Cryzen trading platform is free. This includes 1 hour of back-
tester time per day and 1 live running algorithm at any given time. After the
aforementioned free limits are exhausted, the user is charged a small resource use
fee for additional backtester time or live algorithm deployments (the resource use
fees will typically be a tiny markup on docker container run costs, with the goal
being to simply cover costs).

Further, Cryzen offers premium services for the development of sophisti-

cated external data and machine learning based trading strategies. These premium
services are free to use on the backtester and live for one month. After this one
month trial period, a fixed monthly fee applies to the premium service(s) in use
(exact prices and details for premium service fees and resource use fees are given
in the Cryzen Revenue Sources document). Premium services include:

I Access to twitter data + derived sentiment from tweets

I Access to reddit data + derived sentiment from reddit posts
I Access to google trends data for cryptocurrency related search phrases
I Access to event data, including but not limited to:
• New coin listings
• Network upgrades
• Hard forks
• Exchange hacks
I Access to blockchain / ledger level data, including but not limited to:
• Network speed / confirmation times
• Larger transfers between wallets
• Number of unique wallets
I Access to pretrained predictive machine learning models, including but not
limited to:
• Volatility predictors
• Volume predictors
• Network speed / confirmation time predictors
I Access to raw data1 (1 sec orderbook level data for many cryptocurrencies
across many cryptocurrency exchanges)

More premium services and features will be added to the Cryzen trading
platform over time. Please see Cryzen Revenue Sources document for further detail.

1 The 1 month free access period does not apply to raw data access.
 19 Chapter 4. XEN Token Revenue Sources

4.3 Zen Fund Trading

The Zen fund is an algorithmic trading fund put together by 20% of the capital
raised from the XEN token presale and ICO.

The Zen fund will be traded by algorithms developed by the Cryzen quant
team. The Cryzen quant team will take an AI first approach with initial strategies
including (but not limited to):

I Arbitrage algorithms (enhanced by machine learning models that predict the

length of and volatility during arbitrage windows)
I Momentum algorithms (enhanced by machine learning models that predict
volatility and tendency of prices to move back towards the moving average)
I Social media / sentiment analysis based trading algorithms (algorithms that
quantify the significance and sentiment of news, tweets, and community
voices to identify and act on actionable market insights)
I Google trends based trading algorithms (algorithms that trade based on
Google search data, a window into the collective minds of what people are
considering and searching for at any given time)
I Predictive machine learning based trading algorithms (algorithms that trade
based on machine learning models which try to predict future price changes
given: past price and volume changes, social media sentiment, google search
trends, blockchain / ledger level activity, and more)

Further, the Cryzen quant team will explore the use of reinforcement learning
for trading cryptocurrencies. Please see the Cryzen Revenue Sources document for
further detail into the categories of algorithms that will be used to trade the Zen
fund and projected returns.

As previously mentioned, 40% of all VaaS fees, premium service fees, and
profits from trading the Zen fund will be distributed to XEN token holders.
 T OKEN S ALE

A total of 1,000,000,000 XEN tokens will be minted (no inflationary or deflationary

mechanisms will apply to the supply of XEN token). We intend to raise between $
1 million (soft cap) and $ 2.5 million (hard cap) in ETH from our presale which will
be broken up into two parts: a 1millionraiseviaapubliccrowdsale, anda1.5 million
raise via sales to private (accredited) investors.

Funds raised from our presale will go into finishing the MVP development
of our project (which includes backtesting and live deployment of algorithms
with access to 5s price data feeds) and getting XEN token through the SEC (U.S.
Securities and Exchange Commission) to become one of the first SEC compliant
dividend tokens.

A full ICO will be done some time after our presale (upon compliance with
the SEC). The ICO will be conducted in 3 phases with successively higher prices
per token as described in the diagram below.

A full ICO will be done some time after our presale (upon compliance with
the SEC). The ICO will be conducted in 3 phases with successively higher prices
per token as described in the diagram below.

Over the entire presale and ICO phases, we intend to sell roughly 70% of
XEN tokens. The capital raised therein will be used in the following manner:
21 Chapter 5. Token Sale

Tokens Price per Bonus (USD) Boundsb

for salea 1000 XEN
tokens >100k 10k-100k <10k Upper Lower

SAFT 50 $50 2.5 1.0

ICO 1 200 $125 25% 15% 0% 25.0 18.75
ICO 2 200 $200 25% 15% 0% 40.0 30.0
ICO 3 250 $250 25% 15% 0% 62.5 46.875
∑ 700 130.0 96.625
a in millions
b in million USD

Over the entire presale and ICO phases, we intend to sell roughly 70% of
XEN tokens. The capital raised therein will be used in the following manner:

I Cryzen Trading Platform – 25%

Cryzen trading platform will be allocated 25% of capital raised. This will
be applied towards ongoing platform development, including but not
limited to: supporting more exchanges, collecting data at even higher
frequencies, offering exchange collocated trading services, supporting
more social media outlets and existing social media outlets at even
higher frequency, developing machine learning models that our users
can use out of the box (volatility and price models), and adding raw
22 Chapter 5. Token Sale

machine learning support (so users can train models and optimize their
parameters).
I Zen Fund – 20%
The Zen fund will be composed of 20% of capital raised (to be traded by
the Cryzen quant team).
I Algorithms to Trade Zen Fund – 20%
The Cryzen quant team will be allocated 20% of capital raised. This will
be utilized to develop, test, and optimize profitable trading strategies
for the Zen fund. Please see the XEN Token Revenue Sources - Zen Fund
Trading section for further detail.
I Research and Development – 15%
The R&D division of Cryzen will be allocated 15% of the capital raised.
This will go towards (among other things): the development of rein-
forcement learning support on our trading platform, sophisticated nat-
ural language based detection of major events in the cryptocurrency
space, and building of DAPP (decentralized application) infrastructure
to slowly allow token holders to govern more and more aspects of Cryzen
operations.
I Business Development and Marketing – 15%
Business development and marketing will be allocated 15% of capi-
tal raised. This will be applied towards onboarding new users to the
platform, which will in all probability directly increase the number of
profitable algorithms running on our platform and thus corresponding
VaaS and premium service revenue streams.
I Legal – 5%
Legal costs will be allocated 5% of capital raised to ensure Cryzen and
XEN token are fully compliant with U.S. securities law in both the near
term and long term future.

It is to be emphasized that XEN token is not just another token to be held

and speculated on, but one that is connected to profits of algorithmic trading and
platform services, and therefore has a concrete reason to be valued.
 C RYZEN DAPP V OTING
S YSTEM

Cryzen will also implement a DAPP voting system for token holders to vote on
certain company decisions. Questions will be posed by the Cryzen core team, and
votes will be tallied by a DAPP system where user votes count in proportion to
the amount of tokens held. One of the first use cases of this DAPP voting system
will be to determine what nonprofits our nonprofit fund will donate to (which we
intend to initiate 2-3 years in the future).
 F UTURE V ISION

In this section we speculate on Cryzen and what it may look like a decade into the
future. Token holders will get access to a private Cryzen audit platform, where all
information related to company operations is made available (e.g. anonymized
data related to number of employees, cash spent by category, profits earned, and
more).

Many years into the future, the idea is to turn the entire company into a
DAO (decentralized autonomous organization) which orchestrates the complex
interaction between platform developers, project managers, Cryzen quant team,
and the community to maximize long term profits for token holders.
 A CKNOWLEDGEMENTS

Paddy Lawton – Business Development Adviser

Gareth Fenney – Commercial Adviser
Greg Harris – Business and Community Development Adviser
Shawn Simas – Solidity Smart Contract Developer
Jana Steffen – Whitepaper Editor