A place where businesses could simply hop in to grow their brands.
- Introduction
- The Solution
- What We Solve
- Product Flow
- Why Artificial Super Intelligence [and line of code]
- Why ENS [and line of code]
- Why Kadena [and line of code]
- Additional Features
- GTM
- Roadmap
- Conclusion
So, how do we do it? Well that’s quite simple, one of the methods we use is Loyalty programs, but wait don’t a bunch of businesses have loyalty programs? Let’s tell you why we are different.
Normally, businesses have something known as a loyalty program where a few people enroll to do tasks like joining a discord group, like a few posts on X and so on for a few rewards.
These tasks can get repetitive, people do the same tasks over and over again, which could fail to build actual brand interest, and it needs a community manager too. As you just thought - yes, they don’t grow loyalty in ANY way, nor do they grow the brand.
And most loyalty programs in the web3 space have an EXTREMELY tedious on-boarding phase, which makes it difficult for not so web3 savvy people to be on-boarded.
We saw all these problems and wanted to build something called “BrandX”, an application which could LITERALLY on-board anyone with a simple google sign-in, where a CLUSTER of AI AGENTS, search the ENTIRE internet to identify BOTH the pros and cons of a specific brand.
For every new business on-boarded, they receive an ENS domain which is received by connecting their actual domain, and for every new member who applies for a loyalty program of a specific business, they get a sub name under that ENS domain. This makes it easy to identify brands and users WITHOUT the hassle of going through a series of wallet addresses
In the first phase an agent cluster searches for reviews, comments, news etc and build a KG which consists of the pros and cons of a brand, using which, another agent comes up with a bunch of metrics, quantifying the current condition of a brand with metrics Finally, a bounty agent interacts with the Knowledge graph and with the metrics generation agent to come up with bounties which could boost the areas where the brands lack in. Some of the bounties might include posting an unboxing video, or displaying the camera capabilities of a mobile phone from a specific brand and so on.
Brands review this and finalise their bounties and rewards and submit in the frontend, which uses a business factory contract on Kadena and deploys a contract EXCLUSIVELY for that particular business, and also mints NFTs for bounty completion.
This contract consists of various information such as the bounty details, the prize details, the members enrolled in the loyalty program, the bounties completed by them, the prizes claimable by their members, and so on. This makes sure that no data can secretly be altered by companies.
Rewards would be conditionally issued upon bounty completion, and could be claimed easily by a member of their loyalty program by submitting an image proof of them completing their task, which an agent evaluates and writes on-chain that they have completed that task.
Brands could now focus on ACTUAL customer sentiments and areas where they lack in and grow accordingly.
Loyalty programs can be scammy, bringing it on-chain increases the trust, and brings an on-chain agreement between a brand and its customers that completing a bounty WILL yield them the promised reward.
Unambiguous objectives not aligned with the actual customer sentiment may fail to grow the brand, but this idea causes a COMPLETE paradigm shift in this space where brands could actually grow, WITHOUT a campaign manager/team, reducing more than 20,000 hours/year for brand analysis and growth
- Initiation: A business (e.g.,
domain: business.eth) begins the onboarding process. - Google Sign-in: The business uses Google Sign-in to access the BrandX platform.
- Profile Creation & ENS Verification: The business creates its profile and verifies its domain's ENS (Ethereum Name Service) ownership within the Sepolia test network.
- ENS Registry Interaction: This verification involves a "Check" request to the ENS Registry and a "Verify" response from it, confirming domain ownership.
The BrandX system processes brand data through a sophisticated AI-powered analysis pipeline that transforms raw brand information into actionable insights:
You start by providing brand details (e.g., brand name, identifiers). These details are passed to the Orchestrator, which serves as the central command center for the entire analysis workflow.
The Orchestrator intelligently dispatches tasks to specialized agents. Each agent queries the appropriate MCP (Model Context Protocol) server to fetch relevant information from specific data sources.
a) Web Search Agent
- Searches the web for comprehensive brand-related data
- Passes results into the pipeline for storage in the Knowledge Graph
b) Reviews Agents
- Positive Reviews Agent and Negative Reviews Agent interact with the Reviews MCP Server
- They fetch and filter positive vs negative brand reviews from various platforms
- Results are categorized and stored in the Knowledge Graph
c) Reddit Agents
- Positive Reddit Threads Agent and Negative Reddit Threads Agent interact with the Reddit MCP Server
- They pull brand-related Reddit discussions, classifying them as positive or negative sentiment
- Results are stored in the Knowledge Graph
d) Social Media Agents
- Positive Socials Agent and Negative Socials Agent interact with the Social Media Comments MCP Server
- They fetch brand mentions/comments from social platforms, classifying them as positive or negative
- Results are stored in the Knowledge Graph
All collected and processed results (reviews, Reddit threads, social comments, search results) are stored in the Brand Data Knowledge Graph. This forms a structured and unified data hub about the brand, creating a comprehensive intelligence repository that continuously learns and updates.
The Metrics Generator Agent consumes data from the Knowledge Graph and computes key performance indicators including:
- Sentiment scores
- Volume of mentions
- Ratio of positive to negative signals
- Engagement trends
- Brand health metrics
The Bounty Generation Agent takes the computed metrics via A2A (Agent-to-Agent) communication and generates actionable bounties - specific tasks designed to improve brand performance based on the deep insights gathered from the comprehensive data analysis.
BrandX's customer onboarding is designed to be user-friendly, blending Web2 convenience with Web3 benefits:
- User Initiation: A customer begins their journey on the platform.
- Google Sign-in: The user performs Google sign-in for seamless entry, eliminating complex wallet setups.
- Loyalty Program Request: The user requests to join a business's loyalty program.
- Member ID Assignment: The business adds the user as a member with a unique ENS subdomain ID (e.g.,
user.business.eth). - ENS Subdomain Minting: The business mints an ENS subdomain for the newly onboarded user.
- ENS Registry Update: The subname is written into the ENS Registry, establishing the user's on-chain identity.
- Contract Registration: User details are registered in the smart contract, securing their participation on the blockchain.
Once onboarded, customers can actively participate in bounties and earn rewards through a transparent, verifiable process:
- Bounty Completion & Proof Submission: The user completes a bounty task and submits proof of completion.
- Verification Process: The submitted proof is forwarded to the "Verification Agent" for evaluation.
- Proof Verification: The Verification Agent evaluates the proof:
- Unverified Proof: If verification fails, the user is prompted to resubmit or correct their submission.
- Verified Proof: If successfully verified, the process continues to reward assignment.
- On-Chain Reward Assignment: Upon successful verification, the system adds points and assigns rewards ON CHAIN, ensuring transparent and immutable reward distribution on the blockchain.
We use a cluster of 8 AI agents following the uagents framework, and registered on Agentverse which use 3 MCP servers for brand analysis which perform the following tasks:
- Web Search Agent: Searches the internet for news and product releases from the brand
- Reviews Agents: Look for positive/negative reviews from Yelp, Amazon Reviews, Google Reviews and so on for a product using the Reviews MCP server
- Reddit Agents: Identify positive/negative threads on Reddit for a specific product/brand using the Reddit MCP server
- Socials Agents: Look for positive/negative comments on Social media for a specific brand using a Socials MCP server
All these agents build a Knowledge Graph using meTTa, where we insert various brand sentiments as individual atoms within the KG.
This KG is used by a Metrics Generator Agent which generates 20 extensive metrics like negative_media_coverage_intensity, market_leadership_perception, and so on for a specific brand.
These metrics are sent to the Bounty Generation Agent's address using agent-to-agent communication which finally delivers the necessary bounties for building the brand, and also providing data such as "the points increase in the generated metrics upon a certain number of people completing a specific bounty". This shows that we don't stop at giving personalized bounties, but also provide predictive analysis on areas of growth.
This could help the brands take decisions based on which bounties they want to go with first, depending on the target areas they'd like to improve in. Finally, a Verification Agent verifies the proof submitted by a loyalty program member after completing a bounty and writes the verdict on-chain.
This immediately triggers the reward to be sent to the customer's address if it's an on-chain company, or if it's an off-chain company, customers could claim their off-chain reward by visiting the retailer directly. ASI makes it possible for us to build a customized, objective-driven loyalty program for every brand we on-board and make brands actually focus on growth.
The BrandX platform utilizes a sophisticated multi-agent system built on the uagents framework, with each agent specialized for specific tasks in the brand analysis and loyalty program ecosystem.
- Main Code:
main.py - Purpose: Central coordinator that manages the entire brand research workflow
- Key Features:
- FastAPI-based REST API server
- Sequential coordination of all specialized agents
- MeTTa Knowledge Graph integration for centralized data storage
- Polling mechanisms for asynchronous agent processing
- Brand research status tracking and management
- API Endpoints:
POST /research-brand- Start brand research asynchronouslyGET /research-status- Check research progressPOST /research-brand-sync- Synchronous brand researchPOST /kg/query_brand_data- Query knowledge graph dataGET /kg/get_brand_summary- Get comprehensive brand summaryGET /kg/get_all_brands- List all analyzed brands
- Main Code:
main.py - Purpose: Comprehensive web research using Exa API
- Key Features:
- ASI:One integration for intelligent search decisions
- Exa API integration for web research
- Polling mechanism for long-running searches
- REST API endpoint for brand research requests
- API Endpoints:
POST /research/brand- Perform web research for a brand
- Main Code:
main.py - Purpose: Scrape and analyze positive customer reviews
- Key Features:
- Reviews MCP Server integration
- ASI:One for intelligent tool usage
- Strict formatting for exact review quotes
- Focus on positive sentiment analysis
- API Endpoints:
POST /reviews/positive- Get positive reviews for a brand
- Main Code:
main.py - Purpose: Scrape and analyze negative customer reviews
- Key Features:
- Reviews MCP Server integration
- ASI:One for intelligent tool usage
- Strict formatting for exact negative review quotes
- Focus on areas of customer dissatisfaction
- API Endpoints:
POST /reviews/negative- Get negative reviews for a brand
- Main Code:
main.py - Purpose: Gather positive Reddit discussions about brands
- Key Features:
- Reddit MCP Server integration
- ASI:One for intelligent tool usage
- Single paragraph summaries of positive content
- Community sentiment analysis
- API Endpoints:
POST /reddit/positive- Get positive Reddit posts for a brand
- Main Code:
main.py - Purpose: Gather negative Reddit discussions about brands
- Key Features:
- Reddit MCP Server integration
- ASI:One for intelligent tool usage
- Single paragraph summaries of negative content
- Community complaint analysis
- API Endpoints:
POST /reddit/negative- Get negative Reddit posts for a brand
- Main Code:
main.py - Purpose: Gather positive social media sentiment (Instagram)
- Key Features:
- Socials MCP Server integration
- ASI:One for intelligent tool usage
- Single paragraph summaries of positive social sentiment
- Instagram comment analysis
- API Endpoints:
POST /social/positive- Get positive social media comments for a brand
- Main Code:
main.py - Purpose: Gather negative social media sentiment (Instagram)
- Key Features:
- Socials MCP Server integration
- ASI:One for intelligent tool usage
- Single paragraph summaries of negative social sentiment
- Brand issue identification
- API Endpoints:
POST /social/negative- Get negative social media comments for a brand
- Main Code:
agent.py - Purpose: Generate comprehensive brand metrics and KPIs
- Key Features:
- MeTTa Knowledge Graph integration
- ASI:One AI reasoning for metrics generation
- Comprehensive brand health assessment
- A2A communication with Bounty Generation Agent
- Brand RAG system for data analysis
- Metrics Generated:
- Sentiment metrics (overall, web, reviews, social)
- Reputation risk metrics (crisis severity, vulnerability)
- Market position metrics (competitive advantage, leadership)
- Customer experience metrics (satisfaction, advocacy)
- Performance indicators (health index, resilience)
- Strategic insights (improvement areas, urgency levels)
- API Endpoints:
POST /brand/research- Comprehensive brand researchPOST /brand/query- General brand queriesPOST /brand/data- Specific brand data queriesPOST /brand/summary- Brand summary requestsPOST /brand/metrics- Generate comprehensive metricsGET /brands/all- List all brandsGET /brand/metrics/last- Get last generated metrics
- Main Code:
agent.py - Purpose: Generate actionable bounties based on brand analysis
- Key Features:
- ASI:One AI reasoning for bounty creation
- A2A communication with Metrics Generation Agent
- Brand weakness analysis
- Automatic bounty generation upon receiving metrics
- Creative and actionable bounty suggestions
- Bounty Categories:
- Social Media engagement
- Review generation
- Content creation
- Community building
- Product testing
- Brand advocacy
- API Endpoints:
POST /bounty/generate- Generate bounties for a brandGET /metrics/received- View received metrics dataGET /bounties/auto-generated- Get auto-generated bountiesGET /bounties/auto-generated/{brand_name}- Get bounties for specific brand
- Main Code:
agent.py - Purpose: Verify bounty completion through image analysis
- Key Features:
- OpenAI GPT-4o Vision for image analysis
- Bounty verification system
- Image processing (base64 and URL support)
- Automated task completion verification
- Integration with Bounty Completion Server
- API Endpoints:
POST /chat/image- Image chat with bounty verificationPOST /chat/text- Text chat with optional image
- Main Code:
agent.py - Purpose: Generate positive defense tweets against negative sentiment
- Key Features:
- ASI:One AI reasoning for tweet generation
- Twitter API integration
- Witty and engaging defense responses
- Character limit optimization (<150 characters)
- Automatic Twitter posting
- API Endpoints:
POST /defend- Generate and post defense tweets
- Main Code:
main.py - Purpose: Scrape Reddit posts and discussions
- Key Features:
- Exa API integration for Reddit research
- Multi-mode operation (HTTP API, MCP, or both)
- Comprehensive Reddit post analysis
- Sentiment-specific filtering (positive/negative)
- FastAPI-based HTTP endpoints
- API Endpoints:
POST /scrape-reddit-posts- Scrape Reddit posts with sentiment analysis
- Main Code:
main.py - Purpose: Scrape brand reviews from various platforms
- Key Features:
- Exa API integration for review research
- Multi-mode operation (HTTP API, MCP, or both)
- Comprehensive review analysis
- Sentiment-specific filtering (positive/negative)
- Multiple review platform support
- API Endpoints:
POST /scrape-reviews- Scrape brand reviews with sentiment analysis
- Main Code:
main.py - Purpose: Scrape Instagram comments from brand accounts
- Key Features:
- Apify Client integration for Instagram scraping
- Multi-mode operation (HTTP API, MCP, or both)
- Instagram comment analysis
- Brand-specific account targeting
- Comprehensive social media data extraction
- API Endpoints:
POST /scrape-social-comments- Scrape Instagram comments from brand accounts
ENS serves as the layer of trust and identity in BrandX's loyalty program ecosystem, addressing two critical challenges in the Web3 business landscape:
BrandX leverages ENS's DNS import functionality to establish unbreakable links between businesses and their authentic domain ownership. When a business like Apple.com migrates their existing DNS domain to ENS on-chain, it registers their real domain, against their own wallet address.
This DNS-to-ENS migration process (/src/components/ens-domain-migration.tsx) eliminates the possibility of fraudulent businesses impersonating established brands within our platform. Users can trust that when they interact with apple.com on BrandX, they are genuinely engaging with Apple Inc.'s official loyalty program, not a malicious actor. This verification system builds fundamental trust between consumers, brands, and our platform - essential for any successful loyalty ecosystem.
The on-chain nature of this verification means that once established, the business's authentic identity becomes immutable and globally verifiable, providing unprecedented transparency in brand-consumer relationships.
Every loyalty program member receives a unique ENS subdomain (e.g., sarah.apple.eth) that serves as their permanent, portable identity within the BrandX ecosystem. This goes far beyond simple naming - each subdomain becomes a comprehensive identity profile that bridges multiple data sources seamlessly.
Through our custom offchain resolver (/src/app/api/ens/mint-subdomain/route.ts, /src/app/api/ens/verify-subdomain/route.ts), these subdomains resolve rich metadata including loyalty points from Kadena business contracts, social profiles from our database, and standard ENS records like avatars and descriptions (/src/app/api/ens/fetch-subname-details/route.ts). This creates a unified identity layer where a single ENS name unlocks access to cross-chain loyalty data, social connections, and program benefits.
Members can use their ENS subdomain across any ENS-compatible application while maintaining their BrandX identity and loyalty standing. This portability and interoperability transforms traditional siloed loyalty programs into an open, interconnected ecosystem where member identity transcends individual platforms.
Offchain/L2 Resolver Contract
CCIP-Read Gateway
Documentation
ENS Client API Routes
Kadena is the backbone that makes BrandX possible. Every time a business sets up their loyalty program through BrandX, we use a Business Factory Contract deployed on Kadena to spin up a dedicated smart contract exclusively for that business. This ensures that all bounty details, enrolled members, task completions, and reward claims are transparent, tamper-proof, and verifiable on-chain. It acts as a trust layer of our platform. By deploying each business's loyalty logic as an isolated contract, we guarantee scalability and security without data conflicts.
Our Smart contracts act as a legal agreement between the business and all its loyalty members. Kadena's Chainweb protocol plays a crucial role here. Since Chainweb is a braided, multi-chain architecture where multiple parallel chains work together, BrandX can scale to onboard thousands of businesses without worrying about congestion or high gas fees. Each business contract can execute simultaneously across chains, giving us massive throughput and predictably low costs, which is vital for loyalty programs with high-frequency interactions like bounty submissions and reward claims. It was like Chainweb was tailor-made for this!
On top of that, Kadena's ecosystem provides the NFT infrastructure we use to mint bounty completion tokens for members. These NFTs act as proof of participation and bounty completion, and the contracts also handle token transfers for token airdrop type rewards.
Basically, our platform is a very good use case for Kadena's RWA tokenization goals, by issuing both ERC721 (NFT) with all the voucher metadata for claiming web2 based bounty rewards and also supporting token airdrops for web3 based rewards.
BrandXFactory Contract
BusinessContract Logic
Smart Contract Documentation
Deployed Contract Addresses
-
BrandXFactory Contract
- Explorer Link
- Address:
0x9bb2363810156f7b32b255677e8C1852AC1F95E6
-
Example Business Contract (Apple)
- Explorer Link
- Address:
0xc4ABAd2602D3108371eaA81FdE9A1911b62e4b4a
-
Example NFT Voucher for Loyalty Member
- Explorer Link
- Token ID:
1
BrandX leverages Kadena's robust infrastructure to enable seamless Real World Asset (RWA) tokenization through two primary mechanisms:
When loyalty members complete bounties that offer web2-based rewards (like discounts, free items, or service credits), the system automatically mints ERC721 NFTs that serve as digital vouchers. These voucher NFTs contain rich metadata including:
- Reward details and redemption instructions
- Validity periods and usage terms
- Business branding elements (colors, images)
- Unique identifiers linking to the specific reward template
This approach transforms traditional loyalty rewards into tradeable, verifiable digital assets that members can hold in their wallets as proof of achievement and future value.
For web3-native rewards, the contracts facilitate direct ERC20 token transfers to members' wallets upon bounty completion. This enables businesses to:
- Distribute their own utility tokens as rewards
- Integrate with existing DeFi ecosystems
- Provide liquid, transferable value to loyal customers
- Create token-based economies around their brand
This dual-token approach (NFT + ERC20) is strategically designed to onboard MSMEs (Micro, Small & Medium Enterprises) and corporations into web3 by providing familiar loyalty mechanics with blockchain benefits:
- Familiar UX: Traditional loyalty programs with blockchain transparency
- Gradual Web3 Adoption: Start with simple rewards, evolve to DeFi integration
- Asset Ownership: Members truly own their rewards and achievements
- Interoperability: Rewards can be used across different platforms and businesses
- Programmable Incentives: Smart contract automation reduces operational overhead for businesses
BrandX sells two core outcomes to brands:
A) Continuous Reputation Recovery (signal → action → measurable content)
B) Modern, High-Engagement Loyalty Layer (bounties → token rewards → measurable activation)
Customer Segment (Who to Sell to First):
- Mid-market & large consumer brands in retail / D2C / FMCG / electronics / gaming
- Brands that already run social campaigns and care about UGC
- Prioritizing verticals with frequent physical fulfillment (F&B, lifestyle, D2C) because bounties like unboxings or delivery videos are natural
Positioning:
"Turning reputation issues and low-loyalty activation into measurable growth, where AI finds the problem, prescribes bounties, and issues on-chain, tradable rewards that convert attention into actions."
Primary Buyer Personas:
- Head of Growth / Head of Community / Head of CRM
- CMOs at brand/retail chains
Goals: Running pilots, iterating on verification, capturing case studies, shipping dashboard & analytics.
Deliverables:
- Brand Dashboard: reputation metrics, bounty builder, campaign performance, payout reports
- AI Validation Improvements: video verification model, duplicate detection
- Case Studies from 3 pilots; marketing package
- Basic SDK for integrations + web embed widget
- Billing & Subscription Workflow; basic CSM process
BrandX isn't just another loyalty program platform - it's a complete growth engine for brands. By turning real customer sentiment into measurable insights, generating objective-driven bounties, and bringing transparency through on-chain verification, we eliminate the inefficiencies and trust issues that plague traditional loyalty systems. With the detailed brand analysis from ASI, scalability offered by Kadena's unique Chainweb architecture, and the abstraction offered by ENS, we are able to achieve a user-friendly and powerful brand growth platform.