An anonymous trader from China built an AI system for $120 a month and earned $180,000: a breakdown of the Obsidian + neural network combo
A case has emerged in the crypto community that deserves close attention. An anonymous trader from China, known under the pseudonym CyrilXBT, demonstrated how to create a powerful analytical system at the intersection of the note-taking app Obsidian and artificial intelligence with minimal costs. The result, according to him, is impressive: $180,000 in profit over six months with monthly expenses of only $120.
How does the "smart" setup work?
The system is based on a compact Mac Mini, an iPhone, and local Obsidian storage. Six automated pipelines on the N8N platform operate around the clock: they collect everything the trader reads, listens to in podcasts, and dictates via voice messages in a Telegram bot into the storage. Every night, the neural network scans about 4,000 linked notes, searching for the strongest correlations between fresh information and existing ideas. N8N acts as a link between data sources and storage, automating the entire process without the need for programming.
Every morning at 6:00 AM, a digest arrives by email: three trading ideas with a confidence rating, a forming idea of the week, and any note that contradicts an open position. The system wakes the owner only in two cases—when a new note conflicts with his current thesis or when an idea surpasses the 90% confidence threshold. This allows the trader to avoid distractions from noise and focus only on critically important signals.
The cost and why it matters
CyrilXBT estimates the monthly expenses for running the system at approximately $120—this covers API fees and hosting. The stated profitability is about $30,000 per month, with total profit over six months reaching $180,000. It is worth emphasizing: these figures are not supported by independent evidence, and it is impossible to verify them. However, the example itself is illustrative of a trend.
The author compares such a setup to quantum funds that employ teams of eight people for the same flow of analytical insights. A retail trader achieved a similar result with a single Mac Mini and a monthly budget of $120. This clearly demonstrates how modern technologies—local note-taking apps, AI models, and automation platforms—are blurring the lines between institutional and retail trading.
My expert conclusion: The case is certainly inspiring, but I remind you: large sums on social media are typically not accompanied by evidence. Such systems do not guarantee profits, and their effectiveness heavily depends on the quality of input data and the trader's ability to interpret signals. However, the trend toward personalized AI assistants for trading is unstoppable—and this opens up new opportunities for those willing to experiment.