How I Built a Crypto Scanner That Filters 99% of Solana Noise

jose araujo — Tue, 07 Jul 2026 14:25:50 +0000

Introduction

Every minute, hundreds of new tokens appear on Solana.

Most of them never gain traction.
Some are rug pulls.
Others experience a brief burst of activity before disappearing.

I wanted to answer a simple question:

Can software automatically identify the few tokens that are actually worth watching?

Instead of relying on influencers or manually refreshing DexScreener every few minutes, I started building an automated radar.

The Problem

Most crypto tools rank tokens by volume alone.

The problem is that volume doesn't tell the whole story.

A token can generate thousands of dollars in trading volume while still being extremely risky because:

liquidity is too low
trading activity is artificial
security checks fail
price movement is driven by a handful of wallets

Looking at only one metric creates a lot of false positives.

The Approach

Rather than using a single indicator, I built a scoring engine that combines multiple signals.

Each detected token is evaluated using:

Security validation (RugCheck / GoPlus)
Liquidity
5-minute trading volume
Trading activity
Short-term price momentum

Each category contributes to a score out of 100.

Example:

Security 30 / 30
Liquidity 20 / 25
Volume 16 / 20
Trades 13 / 15
Momentum 10 / 10

Total Score: 89 / 100

The goal isn't to predict winners.

The goal is to filter out obvious low-quality opportunities and surface interesting ones earlier.

Why 5 Minutes?

I experimented with different windows.

One minute creates too much noise.

Fifteen minutes reacts too slowly.

Five minutes turned out to be a good compromise between speed and reliability.

*The Tech Stack
*
The project is built with technologies I use every day.

TypeScript
Node.js
Telegram Bot API
DexScreener API
RugCheck
GoPlus
Cron Jobs

Everything runs automatically without manual intervention.

*Lessons Learned
*
Building the scoring algorithm has been much harder than writing the code itself.

Finding good thresholds requires watching hundreds of real market situations.

Sometimes increasing one score by only a few points completely changes the quality of the alerts.

It's an ongoing process of collecting data, adjusting weights and validating results.

*Current Status
*
The project currently monitors Solana and Base continuously and publishes real-time alerts through Telegram.

My focus now is improving signal quality rather than adding more features.

*Conclusion
*
This project started as a personal experiment to automate something I was doing manually every day.

It has become one of the most interesting side projects I've built because it combines software engineering, data processing and algorithm design.

There is still plenty to improve, but it's already producing useful real-time market intelligence.

If you're building something similar or have ideas for improving scoring systems, I'd love to hear your thoughts.

Crypto Volume Radar is currently available as a free Telegram channel for anyone interested in following the alerts in real time.

DEV Community: jose araujo