Detailed Plan

<u>I N T R O D U C T I O N</u>

<u>A UNIQUE **AI-**DRIVEN MARKETPLACE THAT CAN BUILD AND DELIVER SOLUTIONS ON CARDANO</u>

Imagine a world where creativity rules, formal verification is the law, and an AI-driven marketplace serves up customized applications on-demand. Our Product, QModeler, is built using a unique technology that allows uses to assemble smart contracts like legos. We are ready to take the next step by having AI assemble fully working applications including smart contracts from those Legos.

IF SOFTWARE DEVELOPMENT CAN BE REDUCED AS A FIRST STEP TO THE SAME PROBLEM SPACE AS BUILDING TOYS FROM LEGOS AND ASSEMBLING JIGSAW PUZZLES THEN IT CAN BE AUTOMATED. WE HAVE ACCOMPLISHED THIS FIRST STEP

<https://www.youtube.com/watch?v=TWAD9bVi1YA>

<u>P R O G R E S S T O D A T E</u>

The technology behind QModeler reduces the problem of AI solution development to the problem of assembling toys from legos. QModeler's interactive user guidance feature already uses advanced algorithms to make 90% of the decisions required to build software from specifications. We have prototypes of a working solution - see GitHub. We feel very certain that we are very close to unlocking the key to having AI building software for us from specifications.

<u>U S E O F F U N D S</u>

90% of the funds for this project would be used to host a KAGGLE competition (https://www.kaggle.com/competitions) to attract a large number of teams consisting of acclaimed experts in AI. We may expect to have many teams with experts in several branches of AI competing to produce the wining solution. When the competition is over the solutions of all teams are available to us to select us. The remaining funds will be used to set up and manage the competition.

<u>W H O W E A R E</u>

We believe technology should unite and not divide us. We strive to democratize technology by simplifying it so that innovation happens across all strata of society. Our vision will be realized when any sized company, from enterprise to a small business owner, or even a single person with a yen to create, can innovate solutions to their problems. Ferdon - the maker of Qmodeler is located in Vietnam. QModeler is a supporter of the Vietnam National Computer Olympic Team where many of our team members competed.

Ferdon is a company of currently 20 people passionate about democratizing technology. Ferdon has been working 3 and a half years to build our flagship product, QModeler, one of the first no-code platforms and the first no-code platform for blockchain. QModeler is capable of synthesizing models into smart contracts.

Ferdon is launching the first of its kind, no-code hackathons, long-running events that bring together communities to solve pressing issues using No-Code tools.

Ferdon's active projects include

• In collaboration with WADA we have launched a first of its kind no-code hackathon to develop a solution for food safety traceability in Africa.
• Launching a no-code hackathon in Vietnam to develop a solution to food traceability in Vietnam.

flagship product

Ferdon's flagship product is a low-code platform for developing the blockchain. see link below

<https://youtu.be/njce9Q9hhMc>

<u>A B O U T K A G G L E</u>

Ferdon will be setting up a competent on kaggle to train an AI algorithm to assemble software legos into full applications. Kaggle is the world's largest community of data scientists with:

• 5 million members
• 4 million uploaded solutions
• 300 competitions
• 50,000 open data sets

You can read more about Kaggle competitions here -> https://www.kaggle.com/competitions

<u>A B O U T T H E P R O J E C T</u>

This project will set about to create a new era in software development by introducing AI assembled software. This project is to fund hosting a Kaggle competition (https://www.kaggle.com/competitions) , where we expect several experienced teams to work on completing the prototypes we have developed.

what is graph pebbling

Graph Pebbling is a board game that, because it has a property called Turing Complete, can be used to solve computer problems and is as powerful as any computer language. Graph Pebbling has been the subject of hundreds of research papers.

why is graph pebbling important

An AI solution to create a winning strategy for Graph Pebbling is also a solution to some of the worlds most pressing problems and will unlock a new ways to think about software and software solutions to many of the worlds most urgent problems, available on-demand to those who need them most.

QModeler was built using Graph Pebbling and does not support any other language. Because of this QModeler can be made to synthesize smart contracts, logic for IoT devices, cloud applications,…

how hard is this problem to solve

We are confident it is not hard. The hardest part of solving the many problems is reducing a problem to a know problem space- in our case component assembly. We provide (1) a clear unambiguous definition of how components fit together and (2), a way to unambiguously define what it is we want to be build. (3) a user-friendly way for non-technical users create components and describe what they want to build. (1) and (2) define a problem space with several solutions in solving board games, assembling legos, and solving crossword puzzles.

why use gaggle

There are several reasons why we decided to host a Kaggle Competition (https://www.kaggle.com/competitions)

• access to experts or Kaggle Masters
• training AI algorithms requires expensive specialized hardware
• Kaggle competitions will produce many solutions from many competitors that we want to review.

what kind of applications benefit the most

Any application could benefit but categories that would benefit more than others includes

• applications that involve many parties such as supply chains, food provenance, cold chains, market places
• applications that integrate with other on-chain protocols such as AMM, exchanges, liquidity pools.
• applications that integrate with off applications - oracles, enterprise applications…
• applications that require on-ramping into enterprises such as banks and others
• applications build through no-code crowd sourcing.

<u>M O N I T I Z A T I O N M O D E L</u>

Developers work in a highly centralized model where the middleman, large software vendors, receive the lions share of profits. An AI based system will allow developers to set prices of components they build and track when the component is being used.

We expect around 2% -20% of the revenue produced by an AI generated product will be divided among the builders of the component

<u>I M P A C T</u>

• <u>reduced cost of software</u> - A single toolbox of components can create several, hundreds, of highly customized applications.
• <u>developer productivity</u> - The productivity of developers will skyrocket many fold. Developing a solution will produce components, or legos, that AI can use to create solutions for many problems spaces. AI will take care of assembling the other solutions when they are requested.
• <u>templates creation</u> - The AI marketplace is better than a template, Ai will be able to create several applications from the same legos.
• <u>developer Incentivization</u> - Developers will get continuous streams ofrevenues from components they upload to the marketplace.
• <u>network effects</u> - The project will increase the network effect on Cardano by increasing interoperability. optimistically we could calculate this to be the square if the interoperable components.
• <u>Ecosystem Richness</u> - A toolbox of software building blocks can be the genesis of hundreds, even thousands, of applications.
• <u>More Targeted Applications</u> - Many applications such as food safety have a one-size-fits all approach but sadly that one-size does not fit rural farmers in Africa or Vietnam. Many such applications are highly flows negotiated between a few small farmers and can't be captured in a single application.

<u>F E A S A B I L I T Y</u>

The project will determine if teams on the Kaggle platform (https://www.kaggle.com/competitions) can assemble components built using QModeler where the rules that prescribe which components can be connected and how they should be connected have been unambiguously defined.

• Our team has already achieved over 90% automation ( demonstrated in our video )
• The teams at Kaggle are the best AI experts in the world. The title of Kaggle Master and Kaggle Grand Master is a universally recognized distinction. (https://www.kaggle.com/progression)
• This is a known problem space with several known AI solutions that already exist.
• Often over 1000 teams will compete to to solve a single challenge, with teams using different approaches.
• AI can assemble toys from Legos and solve Jigsaw Puzzles. All these problems have the same characteristics namely (1) a clear unambiguous definition of how components fit together and (2), a way to define what it is we want to be built.
• AI has been highly democratized and with tools like OpenAI that can train algorithm to have winning strategies against board games rather easily. Our technology can be viewed as

<u>R E L E V A N T L I N K S</u>

https://grockious.github.io/lcrl

https://arxiv.org/pdf/2012.11543.pdf

<u>L A U N C H D A T E</u>

November 15 , 2021

<u>P R O J E C T H O M E P A G E</u>

<https://www.cardanonocode.com/projects.html>

<u>B U D G E T</u>

The attached document meetkaggle.pdf has information on the cost

Hosting Kaggle competition including prize –– $200,000

Setting up and management of the competition. –– $20,000

———––

Total –– $220,000