CoraFi
Overview
CoraFi gives token teams one place to deploy and operate Telegram bots across multiple chains. Live buys, sells, burns, and large-wallet activity feed alerts, charts, moderation, scheduled posts, payments, and admin analytics — from a single interface.
Problem
The product receives information continuously and its state never stops changing. Teams were running chain scanners, moderation bots, analytics dashboards, and payment tools as separate products, each with its own state and its own noise.
Most of the design work is status, feedback, and deciding what deserves the user's attention.
Role
Founder. I designed and built the product: architecture, UX, visual design, and the frontend implementation.
System
Ingestion
Live chain events — buys, sells, burns, large-wallet moves — from EVM chains and Solana
Automation & delivery
Telegram Bot API: alerts, scheduled posts, command routing, inline controls
Moderation & analytics
Spam and toxicity flagging, sentiment scoring, engagement dashboards
Access & billing
Subscriptions, wallet payments, promo codes — gated by the same module definitions
Design
Events become actions
Every on-chain event maps to a visible outcome — an alert, a chart, a workflow step. Nothing fires silently, and nothing appears without a source the user can trace.
Human checkpoints on AI
Sentiment and toxicity scoring flag content; moderators keep the controls and the audit trail. The AI narrows attention, it doesn't act alone.
Modules map to plans
The same module definitions drive feature access and billing, so upgrades don't create a second permission system to maintain.
Setup without code
A guided flow — select chains, verify the token, connect Telegram, set branding, deploy. The interface absorbs the complexity so the team doesn't.
Community growth tasks, verified server-side
Moderation suite with AI-assisted flagging
Module-based plans and billing
Implementation
Frontend
Data & events
AI
Billing & delivery
Result
- A live product. Deployment, alerts, moderation, analytics, and billing run from one admin surface.
- Teams go from signup to a deployed, branded bot without writing code.
- On-chain events reach the community as alerts and charts the moment they happen — with the reasoning visible.