Hello everyone,
we’ve been building quietly for a while. It’s time to properly introduce what we’re working on.
We are Sonarlis.
We build infrastructure that detects risk in token launches in real time, from the moment of creation. Not after the rug. Not after liquidity is gone. In the first seconds, before most people have even heard the name.
Most rug pulls are not random. They leave traces early:
- wallet coordination,
- bot-driven buying,
- supply concentration
- deployer history etc..
These patterns exist on chain, but they are not easy to read without the right system behind them. While most tools focus on contracts or static data, we focus on behavior.
Sonarlis monitors blockchain activity from the moment a token is deployed, analyzes behavioral signals across multiple dimensions, and outputs a single composite risk score from 0-100 within seconds. It is not a signal. It is not a recommendation. It is a clear read on what kind of market you are entering, before you decide to enter it.
The Telegram is just the first layer, validation layer. Underneath it, we are building the engine that on-chain markets have been missing.
We have open few free slots on our telegram channel while we fine-tune the engine. Once this phase is over, access will be paid. Early users will have perks later, including lifetime access or discounted pricing.
If you are not convinced yet, follow us and watch how this evolves over the next few days. You might change your mind.
Follow us https://x.com/sonarlis
we’ve been building quietly for a while. It’s time to properly introduce what we’re working on.
We are Sonarlis.
We build infrastructure that detects risk in token launches in real time, from the moment of creation. Not after the rug. Not after liquidity is gone. In the first seconds, before most people have even heard the name.
Most rug pulls are not random. They leave traces early:
- wallet coordination,
- bot-driven buying,
- supply concentration
- deployer history etc..
These patterns exist on chain, but they are not easy to read without the right system behind them. While most tools focus on contracts or static data, we focus on behavior.
Sonarlis monitors blockchain activity from the moment a token is deployed, analyzes behavioral signals across multiple dimensions, and outputs a single composite risk score from 0-100 within seconds. It is not a signal. It is not a recommendation. It is a clear read on what kind of market you are entering, before you decide to enter it.
The Telegram is just the first layer, validation layer. Underneath it, we are building the engine that on-chain markets have been missing.
We have open few free slots on our telegram channel while we fine-tune the engine. Once this phase is over, access will be paid. Early users will have perks later, including lifetime access or discounted pricing.
If you are not convinced yet, follow us and watch how this evolves over the next few days. You might change your mind.
Follow us https://x.com/sonarlis
X (formerly Twitter)
Sonarlis 🟦 (@sonarlis) on X
We’re building a real-time risk infrastructure layer on Base.
• Risk scoring
• Wallet graph analysis
• Liquidity stability modeling
TG https://t.co/G2sMGRQe3c
• Risk scoring
• Wallet graph analysis
• Liquidity stability modeling
TG https://t.co/G2sMGRQe3c
💯2
Sonarlis 🔊
Photo
Every architectural decision in Sonarlis was made with one constraint in mind: the score has to arrive before the trader needs to act.
Everything else is secondary to that.
Speed over completeness. A partial score delivered in 30sec is more valuable than a perfect score delivered in 5min. Every analyzer runs asynchronously with a strict timeout. The pipeline never waits for a slow signal. If one analyzer does not complete in time, the remaining signals are weighted and the score is delivered without it.
Modularity by design. Each analyzer is an independent module. Adding a new signal does not require changing the existing ones. The ingestion layer abstracts chain-specific logic, which means supporting a new chain comes down to implementing a new listener. The analysis engine and scoring logic remain the same.
Calibration before deployment. No parameter goes live without validation. We replay historical token launches through the same pipeline, split the data into training and test sets, and measure precision before anything is exposed to users. The scoring model is not based on assumptions. It is based on observed behavior.
Lean infrastructure.
The current setup runs on:
4 CPUs,
8gb of RAM,
Docker Compose on a Hetzner VPS,
Redis for fast score reads,
and PostgreSQL for persistent data.
Built to be simple now, without limiting how it scales later.
We are not building features. We are building infrastructure. The kind that becomes more valuable the more chains, traders, and protocols depend on it.
Everything else is secondary to that.
Speed over completeness. A partial score delivered in 30sec is more valuable than a perfect score delivered in 5min. Every analyzer runs asynchronously with a strict timeout. The pipeline never waits for a slow signal. If one analyzer does not complete in time, the remaining signals are weighted and the score is delivered without it.
Modularity by design. Each analyzer is an independent module. Adding a new signal does not require changing the existing ones. The ingestion layer abstracts chain-specific logic, which means supporting a new chain comes down to implementing a new listener. The analysis engine and scoring logic remain the same.
Calibration before deployment. No parameter goes live without validation. We replay historical token launches through the same pipeline, split the data into training and test sets, and measure precision before anything is exposed to users. The scoring model is not based on assumptions. It is based on observed behavior.
Lean infrastructure.
The current setup runs on:
4 CPUs,
8gb of RAM,
Docker Compose on a Hetzner VPS,
Redis for fast score reads,
and PostgreSQL for persistent data.
Built to be simple now, without limiting how it scales later.
We are not building features. We are building infrastructure. The kind that becomes more valuable the more chains, traders, and protocols depend on it.
❤3