If you've spent any time trading crypto in the last twelve months, you've probably hit the same wall we did.

The information is there. Hundreds of news sources. On-chain data. Order flow. Macro prints. Dev activity. Twitter. The problem isn't access — it's that no human has the bandwidth to weight all of it, in parallel, every day, without burning out.

So most of us cope by picking a lens. Pure technicals. Pure narratives. Pure macro. We get good at one thing and quietly ignore the rest, hoping the part we ignore won't be the one that takes the trade against us.

xris is a bet that the answer isn't another chatbot. It's a team.

Why one model isn't the answer

When ChatGPT shipped, the obvious idea was: throw the whole job at one frontier LLM. Give it the chart, the news, the fundamentals, ask it what to do.

It doesn't work, and it's worth being precise about why.

A single LLM in a single context window has to flatten everything. It treats a $50M unlock event with the same weight as a Reddit post quoting it. It hallucinates support levels because it pattern-matches on text instead of reading prices. It can't tell you which of its claims came from a primary source and which it inferred. And critically, it has no inner debate — there's no second voice that might disagree with the first.

Markets, on the other hand, are exactly the kind of problem where you want disagreement before you commit capital. The best traders we've worked with don't have one opinion — they have several, weighted, in tension. They know which inputs they trust for what.

What a team of agents actually does differently

xris runs five specialists in parallel:

The S/R agent watches price and volume. It computes support and resistance from a deterministic cascade — anchored to actual touches and volume confirmation, not generated from a paragraph of prose.
The News agent reads hundreds of sources, dedupes, scores sentiment and impact, pins what matters per token. When something hits, it knows it.
The Fundamentals agent tracks supply schedules, FDV/MC overhang, dev vitality, special events — the slow-moving substrate under the chart.
The Macro agent monitors rates, liquidity, calendar events. The exogenous tide that lifts or sinks everything.
The Risk agent is the contrarian. It tracks vol regime, correlation, drawdown sensitivity. Its job is to ask "what if you're wrong?"

Each one is good at one thing. Each one publishes what it sees. A merge layer takes those signals, finds the overlaps, surfaces the conflicts, and presents you with what's left after debate.

You don't get a single answer. You get a prioritized view, with the trail visible.

What surfaces in your dashboard

When you open xris, you see the output of that team:

A levels list with confidence scores and tier annotations — not a vibe.
A news strip per token with what's pinned, why, and from where.
Fundamentals at a glance: supply, vitality, what's coming next.
Live setups — Market entry and Buy-the-dip, recomputed every 60 seconds against the live price using the same ATR cascade we use in backtests.
And when agents disagree, you see it. A bullish technical with a contrary news flag is more useful information than either alone.

That's the point. We're not trying to remove your judgment. We're trying to make sure that when you commit capital, you've heard from every relevant specialist on your bench.

What's next

We're shipping the first production version of xris this month. The five agents above are live. Over the coming releases we'll add agents — on-chain flows, macro narratives, options skew, sector rotation — and deepen the existing ones with more domain expertise per area.

If you want to follow along, the dashboard is open and free during beta. We'll keep posting here as the team grows — including, occasionally, interviews with the agents themselves.

Welcome to xris.