Here's a scenario familiar to anyone who's tried to automate crypto trading: you have a strategy in your head, maybe even sketched out on paper. But turning that into actual executable code? That's where things fall apart. Either you learn Python and wrestle with API documentation, or you settle for a pre-built bot that sort of does what you want, until the market shifts and it doesn't.
Walbi, a Dubai-based platform, is wagering it can collapse that gap entirely. Just describe your strategy out loud—well, in text—and let an AI agent handle the translation.
The company opened its no-code AI trading agents to the public on March 9, following a 14-week closed beta that generated some eye-catching numbers: more than 1,000 participants built over 9,500 agents, which collectively executed 187,000 autonomous trades. The premise is deceptively simple. You tell the system what you want—timeframes, risk thresholds, when to enter, when to bail—and it figures out the rest.
Whether that simplicity holds up under market pressure is another question entirely.
Trading strategies, no assembly required
At the core of Walbi's approach is natural language processing. Users spell out a trading strategy conversationally, and the platform's AI—powered by OpenAI's API—converts that into executable logic. The agents operate inside Walbi's environment, drawing on portfolio data, technical indicators, an economic calendar, the Crypto Fear & Greed Index, and liquidation data to inform their moves.
"Retail traders are familiar with automation, but most existing tools require coding or rigid rule configuration," Anthony Cerullo, the company's Chief Commercial Officer, said in the launch announcement. "Our goal with AI agents is to make strategy automation accessible through dialogue, while maintaining transparency and user control."
There's no scripting language to master. No flowcharts to construct. The workflow: create an account, articulate or digitize a strategy, backtest it, tweak the risk settings, deploy it live.
Easy enough, in theory.
What a beta actually tests
The closed beta ran from October 2025 through January 2026, giving Walbi a sandbox to stress-test before going wide. Performance was, predictably, all over the map. That's trading. Market conditions shifted. User configurations varied. According to the company, momentum-oriented strategies that incorporated Fear & Greed data and liquidation metrics held up best during volatile stretches.
But the beta wasn't designed to answer whether these agents can consistently beat the market—a claim Walbi wisely avoids making. Instead, the company focused on the less glamorous but more critical questions: stability, execution reliability, risk controls. The infrastructure concerns that matter when automation is managing actual capital. Over 187,000 trades across thousands of agents gave them enough data to surface edge cases and refine the system before opening the floodgates.
Still, the real test comes when users risk their own money. Beta performance is one thing. Live markets under stress are something else.
A marketplace for trading logic
Perhaps the more intriguing piece of the launch is Walbi's Agent Marketplace, where users can list their agents publicly. Each listing displays transparent performance history, drawdown data, strategy type, follower count. Other traders can browse, compare, and deploy someone else's creation. And crucially—creators can monetize their strategies, turning profitable trading logic into a revenue stream.
Walbi says it won't charge commissions for agent creation, positioning itself as infrastructure rather than taking a direct cut of performance.
It's a model that aligns incentives in an interesting, if uncertain, way. Successful creators earn income. Users gain access to battle-tested approaches. Walbi grows its ecosystem. Whether this creates a marketplace of genuinely valuable strategies or devolves into a showcase of overfitted backtests that looked great in 2025 but crumble in 2026? That remains very much to be seen.
Marketplaces have a way of surfacing quality over time—or drowning it in noise.
A landscape getting more crowded by the month
Walbi isn't the only player chasing this vision. The retail crypto trading automation space has gotten increasingly competitive. Platforms like Coinrule offer IFTTT-style no-code builders. Bitsgap provides DCA, grid, and futures bots with AI assistance. Kryll has offered a drag-and-drop editor and marketplace model for years.
More recently, the shift toward truly autonomous agents has picked up steam. Senpi launched personal trading agents for Hyperliquid in late February, also using a no-code setup. Around the same time, 0x rolled out a cross-chain API beta designed to power "agentic swaps" across more than 15 blockchains, while XRPL introduced an x402 facilitator for AI agent payments.
Infrastructure is standardizing around a new assumption: agents, not just humans, will be executing trades. Increasingly, that assumption looks reasonable.
Walbi's bet on adaptive agents rather than static bots is meant to carve out differentiation. The emphasis on real-time data ingestion and dynamic decision-making is supposed to separate it from older rule-based systems. Whether traders actually care about that distinction will depend on one thing. Returns.
The fine print you probably should read
Walbi operates as Walbi LLC, registered in Saint Vincent and the Grenadines—a jurisdiction popular among crypto platforms, though not exactly renowned for rigorous regulatory oversight. The company reports 2.9 million registered users as of March, a significant jump from earlier figures that hovered between 1 million and 1.7 million through 2024 and early 2025. No independent audit of those numbers exists.
The platform is available on iOS and Android, though geo-restriction messages have reportedly appeared for users in certain regions. Walbi offers leverage up to 500x, a feature that underscores the high-risk nature of the environment, regardless of how user-friendly the interface becomes.
Anecdotal complaints about fees and trading losses have surfaced on Reddit and Trustpilot. Unverified, yes. Not necessarily representative of broader user experience, sure. But worth noting all the same. Making strategy automation easier doesn't make trading itself less risky.
The question Walbi is betting on: does accessibility alone unlock a new tier of retail participation? If describing a strategy in plain English lowers the barrier enough, and if the marketplace surfaces genuinely useful agents rather than just convincing-looking junk, the platform could establish a foothold in a space where many have tried and stumbled.
But the crypto trading graveyard is littered with tools that promised to simplify the complex. Whether Walbi's agents prove durable or disposable will come down to performance. In trading, that's the only metric that ultimately matters.
