Created on 15th May 2025
•
[Rebranding X-Alpha to RogerAI]
Roger AI is the AI and data backbone for web3 InfoFi—bringing social and on-chain intelligence together to power smarter web3 experiences.
The core problem we’re solving is data fragmentation in web3. Existing platforms either focus on social signals or on-chain activity, but none unify both. This siloed approach makes it difficult for traders, analysts, and applications to make real-time, informed decisions.
Roger AI bridges this gap by serving as a unified AI and data layer that sits on top of both blockchain and social media data (e.g., X/Twitter, Telegram, Farcaster). It transforms this raw, noisy data into actionable insights—enabling applications ranging from research tools and co-pilots to trading assistants, screeners, and alert systems. We have several pipelines to extract useful information like narrative, events, sentiment etc from the social data
We’ve built and shipped multiple products on this foundation:
With over 40,000 users and $100K in revenue in the last year from ads and subscriptions, Roger AI is already proving product-market fit.
As the demand for AI agents, trading co-pilots, and smart discovery tools grows, we aim to be the go-to data infrastructure powering these next-gen InfoFi experiences.
About XAlpha
Retail crypto traders are overwhelmed by noise and lack actionable intelligence. While platforms like Twitter, Telegram, and on-chain explorers are rich in alpha, most users don’t have the tools to filter signal from noise, identify opportunities early, or take action instantly. Existing solutions either focus on price charts or general analytics—none combine real-time social intelligence with onchain execution in one seamless flow.
X-Alpha.ai bridges the gap between insight and action.
We’ve integrated Coinbase’s AgentKit to allow users to perform onchain actions (like buy/sell or approve) directly from within our platform. But we go a step further—we combine it with X-Alpha’s proprietary social analytics engine to surface the why behind token movements.
Our AI Assistant now does 3 key things:
Surfaces early signals using Twitter and Telegram data: trending tokens, KOL mentions, follower spikes, and narrative heatmaps.
Guides user decisions through conversational insights, helping them understand social momentum and risk signals before acting.
Enables instant execution via Coinbase’s AgentKit—no need to switch tools or miss timing.
This unlocks a first-of-its-kind trading flow:
Discover → Validate → Execute — all within one unified interface.
By combining context, clarity, and execution, we solve the core UX pain of crypto trading today.
There are several challenges/obstacles that we ran into:
Integrating Coinbase AgentKit with our existing backend
Coinbase’s AgentKit is powerful but designed with a specific structure in mind. Our backend (Python/Flask) had to be extended and adapted to handle wallet delegation, action simulation, and secure execution without compromising our existing stack.
We overcame this by modularizing the integration and extending AgentKit's action provider with several other tools (including the Xalpha APIs/MPCs) along with Langgraph integration for graph based workflow.
Mapping social insights to onchain actions
X-Alpha’s intelligence layer deals primarily with social signals (mentions, trends, influencer activity), while AgentKit expects clear onchain intents (buy, sell, swap).
Bridging this gap was non-trivial—we had to create an AI-driven decision layer that translates social alpha into actionable trading suggestions/actions.
Tracks Applied (5)
Technologies used
Cheering for a project means supporting a project you like with as little as 0.0025 ETH. Right now, you can Cheer using ETH on Arbitrum, Optimism and Base.