Hi, I'm Alex.
- Web3Fuel Founder
I started Web3Fuel after one too many "this AI demo will never survive contact with real markets" moments.
As someone trading and researching markets daily, I was tired of seeing AI tools that worked great in screenshots but couldn't be trusted with money on the line. So I decided to build the systems I wished existed.
Turns out, the gap isn't AI capability — it's calibration. Most AI tools never track their own accuracy. Most never close the loop between prediction and outcome. And most documentation assumes you already know what they're missing.
Web3Fuel exists to build AI systems financial teams can actually trust.
Through prompt engineering, calibrated feedback loops, and production tooling, I ship the kind of AI workflows that close the gap between "demo" and "real money."
These aren't backtests or cherry-picked examples. Every prediction is tracked against live market data, scored honestly, and fed back into the model to improve future calls. That's what production AI actually means.
Lately the work has moved toward AI-driven market analysis tools.
The Macro Scanner is the main one — an AI pipeline that ingests macro news, runs deep-dive research on the highest-impact stories, and produces specific ticker trade calls with entry/target/stop. It's been running continuously since February 2026 with 86% directional accuracy across 144 resolved trades, and 94% on highest-confidence calls.
Around it: a Polymarket monitor that flags unusual liquidity spikes on prediction markets, and a Discord bot framework that ties the alerts together. Every prediction is tracked, scored against actual market data, and the feedback loop is wired into the model's prompt to calibrate future calls.
Open to roles where AI-driven analysis meets real-world decisions — particularly in financial markets, asset management, or any space where calibrated AI signals create measurable value.
More tools, deeper analysis, broader scope.
The macro scanner will keep growing — better calibration, larger sample size, more transparent methodology. Adjacent: a content automation engine that turns scanner output into platform-specific social posts with a human approval layer, and analyst workflow tools that compress research cycles from days to hours.
Whether the next chapter is AI for markets, AI for marketing, or AI for research operations — the principles stay the same. Track what you predict, score honestly, learn from outcomes.
Whether you're working on AI-driven analysis, marketing automation, research workflow tools, or want to chat about how to build systems that learn from their own predictions — I'd love to hear from you.
Get In Touch