An intelligent solution to analyze NFT collectibles, detect fraudulent activities and minimize risks
Vasiliy Karpitski and Alexei Dulub are entrepreneurs and blockchain enthusiasts with more than 30 years of joint experience in technology and software development. After analyzing the market potential as well as challenges associated with NFT investment, the technology evangelists decided to join hands to develop a smart solution that would help NFT investors make informed decisions.
The PixelPlex team has delivered a smart platform that provides users with actionable data on NFT collectibles, their provenance and ownership, thus enabling NFT enthusiasts to reduce risks and earn revenue. It also addresses the needs of businesses and analytics firms that can quickly access a large array of structured data and the results generated by ML models for analytical purposes.
The exponential growth of NFTs has attracted many creators, businesses, and NFT enthusiasts, willing to invest in the NFT market. In 2021, the sales of NFTs hit $17,6 billion with most of them occurring on secondary markets. NFT collectibles have turned into one of the most profitable capital investments. At the same time, the huge growth of the NFT sphere has eventually resulted in an increasing number of scams and frauds like blacklists, wash trades, duplicates, and more. Therefore, NFT buyers, creators, and businesses are now struggling to make quality investment decisions and avoid risks.
To help NFT enthusiasts and investors avoid fraud, evaluate risks, make better decisions, and earn yield, the startup decided to build CheckNFT.iO — a platform that would collect and analyze large amounts of data from the blockchain network and ML models. The platform would allow anyone interested in the NFT industry to easily verify NFT collectibles, find gems, and tackle scams while delivering tangible benefits for enterprise customers and analytical firms by providing them with easy access to up-to-date NFT data. The PixelPlex team was assigned to build the CheckNFT.iO solution and accomplish the following goals:
Create a platform that would allow users to browse, compare, and analyze data behind NFT collectibles
Equip the solution with an AI engine that analyzes NFT’s history on blockchain and monitors newly created blocks in real time to provide users with potential copies or IP exploits
Add advanced analytics tools like marketplace analytics, top NFT gems, etc.
Prevent scams, blacklists, price pumps, wash trading, and other financial manipulations with the help of AI algorithm that detects the fraudulent smart contract creations, mints, and other suspicious activities, thus enabling users to invest in a much safer way
Ensure intuitive navigation so that both NFT newbies and professionals in the field could easily use the service
Sophisticated, intelligent platform that pulls data from multiple sources (blockchain, marketplace, metadata standards, etc.) for NFT analysis
Data intelligence and analytics tools that leverage AI-enriched data to track trends and deliver actionable data
Risk analysis tools that make use of AI to evaluate various scenarios like the ability to change the number of tokens or metadata in a project
Trained and deployed ML models that find fakes through image and text analysis and help users prevent losses
Browser extension for top marketplaces like OpenSea, Rarible, SuperRare, Nifty Gateway, MakersPlace, and Foundation to ensure seamless user experience
Intuitive UI/UX for users of all levels
The largest database of tokens and NFT collections
NFT performance indicators and value projection forecasts
AI-based NFT analysis
ML models to compare NFTs and find fakes
Risk alerts and notifications on fraudulent activities
NFT side-by-side comparison
Support of top existing marketplaces
Statistics on top NFT gems and collectibles
Got an idea? Let’s work together
To help users avoid fraud and scams, we implemented a risk analysis module that provides a summary on critical, middle, and minor risks. The module applies the AI engine to detect fraudulent and suspicious activities and uses ML models to detect scams and find fakes through image analysis. CheckNFT.iO contains the following risk parameters:
To help users avoid trading malpractices like blacklists, wash trades, and price pumps, the platform will also inform them on price manipulation, dump sale, IPR infringement, and duplicates risks.
Our team equipped the CheckNFT.iO service with data intelligence and artificial intelligence algorithms that benefit both NFT artists and businesses. The service continually evaluates NFT’s history on blockchain and monitors newly created blocks in real time, thus providing users with actionable data. We also built the functionality that allows the platform to send IP violation notifications to NFT artists and project creators, thus helping them protect their artwork from fraud. The service is also beneficial for businesses touching NFTs by providing them with alerts on fraud detection, wallets analysis and profiling, and more.
To help businesses and NFT enthusiasts get a complete picture of each collectible at a glance, the data intelligence service collects information from multiple sources, including blockchain, metadata standards, and marketplace.
Blockchain
By literally decomposing each block into atoms through a custom-built Ethereum node, the platform receives data on token basic information, token creation transaction, history of token transfers, and token metadata address.
Metadata standards
The platform collects such metadata as token name and description, token URL, and additional attributes from multiple sources while being able to pull data from broken links through its IPFS service.
Marketplace
Token listing price is the only information that is requested directly from a marketplace as it is not stored on the blockchain. We stick to the immutable source of truth, therefore the platform collects all other information from the blockchain network.
To empower NFT enthusiasts to easily find most prominent NFTs, the platform’s ML algorithm forms Top Collectible Sales along with gems from Top Volume Collections, which can be filtered by day, week, and month. The platform also showcases the following useful stats:
Got an idea? Let’s work together
CheckNFT.iO is built on top of the Nest.js and Next.js frameworks leveraging TypeScript, JavaScript, and Go. To ensure increased scalability and flexibility, the platform is built using a microservices architecture and ETL process. As a result, adding new marketplaces, blockchains or preparing data for ML models does not require making any changes to the core of the platform, which significantly reduces development time.
The CheckNFT.iO platform comprises the following system components:
API service
The service is responsible for handling API requests.
Token processor
The service connects CheckNFT.iO with CryptoAPI on top of which the platform is built. The CryptoAPI is responsible for working with blockchain raw data, which is further processed and enriched by the CheckNFT.iO platform.
BC Connectors
Each blockchain has its separate blockchain connection process.
MP Connectors
Each marketplace has its separate blockchain connection process, which provides for the unification of requests and data processing.
MetaData Connector
The service responsible for handling data on the contract’s metadata was also designed as a separate module to unload the processes of the main services.
BC Watchers
The service is responsible for finding new contracts. It monitors new blocks, identifies transactions in which a new contract was created, and tries to determine whether the contract is an NFT contract.
MP Watchers
Each marketplace should have a separate service launched. The service is responsible for detecting new events and, if such found, informs the Token processor about them.
CheckNFT.iO also applies ML models that continuously process input data while being retrained and optimized. Currently, the platform uses 3 different ML models, which:
Compare the names of the tokens to ensure that there are no manipulations with symbols or replaced letters
Search for similar images by building embeddings with convolutional neural networks and then finding the closest image using the nearest neighbor algorithm
Analyze and detect wash trading activities by connecting interacting groups and identifying unusual patterns
Within the first month since its launch (as of June 2022), CheckNFT.iO boasts the following results:
71M
tokens are processed on Ethereum
134M+
events processed (sales, transfer, minting)
176k+
NFT collections in total