The problem CredBuzz solves

At CredBuzz, we’re building Infofi — an AI-powered, decentralized platform that connects brands with highly relevant influencers for marketing campaigns.

Our AI matchmaking system identifies niche creators whose content aligns with a brand’s target audience, driving higher user growth and brand resonance.

We use ML and large language models (LLMs) to semantically analyze content, trends, and creator behavior. By deeply understanding the quality, and influence of each creator's content, we assign engagement scores and label creators based on interest, expertise, and content quality. This enables two core use cases:

1. Open Campaigns – Brands can launch decentralized campaigns inviting public participation. Our smart contract–based escrow system distributes rewards trustlessly and transparently based on actual performance, measured through our AI engagement scoring.

2. Influencer Discovery – Our AI matchmaking system identifies niche creators whose content aligns with a brand’s target audience, driving higher user growth and brand resonance.

Additional features include:

• Smart-contract driven partnerships
• Automated endorsement verification
• Performance-based payment models
• Transparent and secure transactions
• Deep campaign analytics

We aim to make influencer marketing scalable, trustworthy, and efficient by leveraging AI and decentralization.

Current Focus:
For our MVP, we are targeting the crypto/Web3 market, where incentive-driven marketing is already well established especially on platforms like X (formerly Twitter). In this space, brands and creators commonly use financial rewards and airdrops to drive engagement, making it a natural entry point for CredBuzz.

Challenges I ran into

Following are the major obstacles encountered through the dev journey of this product:

Onchain Challenges

1. Smart Contract Security
   Handling stablecoin escrows and payouts safely is critical. We addressed this with rigorous testing, and got the contract audited by third-party experts.

2. Onchain endorsement verification
   Social engagement is offchain by nature. We tried using oracles to feed verified performance data into smart contracts, to preserve decentralization. But it turned out to be much more complex than anticipated, so currently endorsement verification is web2 way only.

Offchain Challenges:

1. Data Aggregation from Social Platforms
   APIs can be limited or rate-restricted. this was mitigated with compliant scraping and intelligent caching.

2. Real-Time Semantic Analysis at Scale
   Processing and analyzing millions of tweets is compute-intensive. For company like mine, which is bootstrapped, tried innovative ways of using free api credits from different LLMs (creating multiple accounts + excelling at prompt engineering)