ZoroX exists because of a complaint every memecoin trader knows by heart: by the time a token shows up in a Crypto Twitter thread or on an aggregator’s trending list, the move is usually over. The early buyers are already taking profit, and everyone else is buying their exit liquidity.
ZoroX was built to attack exactly that timing problem. It is an autonomous AI agent running on Solana that watches TikTok twenty-four hours a day, looking for memecoins that are starting to catch fire — before that attention spills over into the wider market. The project was built by Gabriel Antony Xaviour and first surfaced through the hackathon ecosystem, appearing on DoraHacks as a working proof-of-concept that wires social-media virality signals directly into on-chain trading data.
This guide walks through what ZoroX actually does, the mechanics underneath it, who might find it useful, and — just as important — where its limits are. No hype, no price predictions, just a clear-eyed look at the tool.
What Is ZoroX?
The short answer: ZoroX is an AI-powered monitoring agent on the Solana blockchain that continuously scans TikTok for crypto-related content, identifies memecoin tickers gaining viral traction, and matches them against live on-chain token data.
Think of it less like a trading bot and more like a tireless research assistant. It does not sleep, it does not get distracted, and it does not miss a video because it was scrolling something else. Its whole job is to sit at the intersection of two worlds that move at very different speeds: TikTok’s chaotic, minute-by-minute content churn and Solana’s on-chain token markets.
The insight behind the project is straightforward. TikTok’s short-video format has become one of the fastest places on the internet for retail crypto interest to surface. A coin can rack up thousands of mentions in video descriptions and comment sections hours — sometimes days — before that same interest shows up in trading-focused communities. ZoroX exists to catch that window.
It is worth being clear about what ZoroX is not. It is not an exchange, not a wallet, and not a signal service telling you what to buy. It is an analytics and detection layer: it finds the signal, structures it, and connects it to verifiable on-chain data so a human can decide what to do with it.
How ZoroX Works
The system runs as a pipeline with three main stages. Each one handles a different part of the problem, from raw noise to structured, tradable information.
1. Round-the-Clock TikTok Scanning
ZoroX continuously scans TikTok for posts carrying cryptocurrency-relevant hashtags — tags like #solana, #pump, #pumpfun, #meme, and #memecoin. It works in rolling 24-hour windows, which means it is always looking at the freshest slice of content rather than a stale snapshot.
This matters more than it sounds. Virality on TikTok is brutally time-sensitive. A token that gets mentioned in fifty videos on Tuesday morning might be old news by Tuesday night. The rolling-window design is the project’s answer to that: always fresh, always moving.
2. AI Classification and Ticker Extraction
Raw hashtag matches are noisy. Plenty of videos use #meme without having anything to do with crypto, and plenty of crypto videos are jokes, scams, or reposts. This is where the AI layer earns its keep.
ZoroX uses OpenAI’s language models to do two things. First, it classifies whether a piece of content is genuinely crypto-related or just riding a hashtag. Second, it reads video descriptions and comment sections to extract ticker symbols — the actual coin names people are talking about.
This is harder than it looks. Tickers get misspelled, abbreviated, or buried in slang. Comments are full of noise. Getting a clean list of “which coins are actually being discussed” out of that mess is the core technical challenge, and it is the part of the system where the AI models do the heaviest lifting.
3. Cross-Referencing With Pump.fun Token Data
A ticker name alone is not enough to act on — especially in the memecoin world, where five different tokens can share the same name and four of them are scams. So ZoroX takes the extracted tickers and cross-references them against Pump.fun’s token database, pulling back contract addresses and on-chain trading data.
This step is what separates a fun social-listening toy from something a trader can actually use. The contract address is the ground truth: it tells you exactly which token the chatter is about, lets you check its liquidity, holder distribution, and trading history, and filters out the copycats. Without it, you are just reading vibes. With it, you have something verifiable.
Key Features of ZoroX
Pulling the pieces together, here is what the system actually delivers to someone using it:
- 24/7 autonomous monitoring. The agent runs without human intervention, watching TikTok in rolling 24-hour windows so nothing in the current cycle slips past.
- Hashtag-targeted scanning. It focuses on the tags where Solana memecoin conversation actually happens — #solana, #pump, #pumpfun, #meme, #memecoin — rather than trying to boil the ocean.
- AI content classification. OpenAI models filter out irrelevant content and confirm which posts are genuinely about crypto before any ticker is extracted.
- Ticker extraction from descriptions and comments. The system reads where the real conversation happens — not just video captions, but the comment sections underneath them.
- Pump.fun database matching. Identified tickers are matched to real tokens with contract addresses, cutting through the duplicate-name problem that plagues memecoin trading.
- On-chain data linkage. Once a token is identified, its live trading data is attached, so social buzz can be weighed against actual market activity.
- Early-signal positioning. The whole design is oriented around one thing: surfacing tokens while the viral traction is building, not after it has peaked.
Who Is ZoroX For?
Not everyone needs a tool like this. But for a few specific groups, it fills a genuine gap.
Active memecoin traders are the obvious audience. If your strategy depends on catching narratives early, a system that watches TikTok so you do not have to is directly useful — especially on Solana, where the memecoin cycle moves fast enough that hours matter.
Market researchers and analysts get a different kind of value. ZoroX maintains a structured record of which tokens trended, when, and alongside what kind of content. Over time, that becomes a dataset for studying how social virality translates into market movement — useful for anyone writing about or modeling crypto attention cycles.
Content creators in the crypto space can use it in reverse: seeing which narratives are gaining traction helps them understand what their audience is already curious about.
Developers and hackathon builders might find the architecture itself interesting. ZoroX is a clean example of an agentic AI system with a real data pipeline — social ingestion, LLM classification, on-chain verification — and it emerged from exactly the kind of experimental environment where those designs get tested first.
ZoroX vs Traditional Memecoin Research
To put the tool in context, it helps to compare it against how traders usually hunt for early memecoins:
| Manual research | ZoroX | |
|---|---|---|
| Coverage | A few hours of scrolling per day | 24/7 autonomous scanning |
| Speed | Finds trends after they are visible | Rolling 24-hour windows catch trends as they build |
| Noise filtering | Human judgment, inconsistent | AI classification of every post |
| Token verification | Manual contract lookup | Automatic Pump.fun cross-reference |
| Scalability | One person, limited attention | Thousands of posts processed continuously |
| Cost | Time | Compute and API usage |
Manual research has one advantage the table does not show: human intuition. An experienced trader can smell a coordinated shill campaign or recognize a joke that an AI might take seriously. ZoroX is a complement to that judgment, not a replacement for it.
Risks and Limitations
Any honest guide to a tool like this has to talk about where it falls short. Here are the main ones.
Social signals are gameable. TikTok engagement can be bought, botted, or coordinated. A token can look “viral” because a marketing team paid for it to look viral. ZoroX detects traction; it does not verify that the traction is organic. That distinction matters enormously.
Correlation is not causation. A coin trending on TikTok does not mean it will perform well, or perform at all. Social buzz and price action are related but separate things, and plenty of viral tokens go nowhere — or worse.
AI classification is imperfect. Language models misread sarcasm, miss context, and occasionally extract the wrong ticker. In a market where acting on the wrong contract address can mean buying a scam token, that error rate is not a footnote.
The memecoin market is inherently high-risk. This is the big one. Most memecoins lose most of their value. Tools that help you find them faster do not change that base rate — they just help you find them faster. Anyone using social-signal tools should size positions accordingly and never trade money they cannot afford to lose.
It is still early-stage. ZoroX emerged from the hackathon ecosystem as a proof-of-concept. That is a strength in terms of innovation, but it also means users should treat it as experimental technology rather than battle-tested infrastructure.
Frequently Asked Questions
What is ZoroX in simple terms?
ZoroX is an autonomous AI agent built on Solana that watches TikTok around the clock for memecoins gaining viral attention. It scans crypto-related hashtags, uses AI to identify which coins people are actually talking about, and links those coins to real on-chain token data via Pump.fun.
Who built ZoroX?
ZoroX was built by Gabriel Antony Xaviour. The project emerged from the hackathon ecosystem and surfaced on DoraHacks as a functional proof-of-concept connecting social media virality signals to on-chain trading data.
Which blockchain does ZoroX run on?
ZoroX is built on Solana. That choice makes sense given the project focuses on Solana memecoins and the Pump.fun ecosystem, both of which live on Solana’s fast, low-cost network.
How does ZoroX find memecoins on TikTok?
It continuously scans TikTok for posts using crypto-relevant hashtags like #solana, #pump, #pumpfun, #meme, and #memecoin within rolling 24-hour windows. OpenAI language models then classify whether the content is genuinely crypto-related and extract ticker symbols from video descriptions and comments.
What does ZoroX do with the tickers it finds?
It cross-references them against Pump.fun’s token database to retrieve contract addresses and on-chain trading data. This step matters because memecoin names are often duplicated — the contract address confirms exactly which token the social chatter refers to.
Is ZoroX a trading bot? Can it buy tokens for me?
No. ZoroX is a detection and analytics layer, not a trading bot. It finds and structures early signals; it does not execute trades or tell you what to buy. Any trading decisions remain entirely with the user.
What are the main risks of relying on a tool like ZoroX?
The biggest risks are that social signals can be faked or coordinated, AI classification can make mistakes, and viral attention does not guarantee price performance. Memecoins as a category are extremely high-risk, and faster discovery does not change that. Treat ZoroX as one input among many, not as a decision-maker.
Conclusion
ZoroX is a genuinely interesting answer to a real problem. Memecoin traders live and die on timing, TikTok has become one of the earliest places retail crypto interest appears, and nobody can watch the platform manually at the scale an AI agent can. By chaining TikTok scanning, AI classification, and Pump.fun verification into one autonomous pipeline, ZoroX compresses hours of research into a structured early signal.
But a signal is all it is. Social traction can be manufactured, AI makes mistakes, and the memecoin market remains one of the most unforgiving corners of crypto. Used with clear eyes — as a discovery aid alongside real due diligence — ZoroX is a smart piece of tooling. Used as an oracle, it will eventually teach an expensive lesson. The traders who get the most out of it will be the ones who understand both what it sees and what it cannot.
