Built by traders who got tired of losing edge to bad infrastructure

The market we set out to fix

By 2023 it was clear that the most active, profitable corner of crypto trading had drifted onto Solana, into memecoins, perpetuals, and a handful of fast-moving DeFi venues. It was also clear that the tools serving this market were either retail-grade wallets bolted onto aggregators or institutional terminals priced for hedge funds. There was nothing in the middle that respected both the user's keys and the user's clock.

The opportunity was specific: build a single non-custodial venue that combined discovery, execution, and analytics, with latency comparable to a top-tier centralized exchange and a UX that didn't require a quant background to use safely.

The first six months

We started with execution because execution is where users get extracted. A team of three built the original transaction router that bundled Jupiter, Raydium, and direct pool routes behind a single quote, with MEV protection as a first-class option rather than a checkbox buried in advanced settings. That router still powers every spot trade on Axiom today, hardened by tens of millions of fills.

The Pulse insight

Discovery was the harder problem. Existing tools showed *what had happened*; traders needed *what was happening*. Pulse — three parallel streams of New Pairs, Final Stretch, and Migrated tokens — emerged from a simple observation: the best memecoin traders we knew were running three browser tabs and a Telegram bot to manually approximate this view. We built it natively, with sub-second event ingestion and filters that actually work.

What we believe

We believe non-custodial is not a feature to apologize for; it is the only acceptable default. We believe latency is a UX feature, not an engineering one. We believe transparency about fees, MEV, and slippage builds long-term trust in a market that has trained users to expect the opposite. And we believe the people building this should be visible — the team page is a real list of people, not stock photos.

The road ahead

Spot and perps are the foundation. The roadmap extends into structured products built on the same execution layer: cross-venue arbitrage primitives, programmatic strategies you can deploy without writing code, and an API that lets other teams build on top of the same routing and protection guarantees. The throughline is the same as day one — give serious traders serious tools, and stay out of their way.

Strategy in a sentence

Compound the platform's structural advantages — non-custodial custody, sub-second execution, integrated MEV protection, unified discovery — by deepening each one and connecting them more tightly to the workflows of serious traders. Expand selectively where new market structures (cross-venue, programmatic) reward the same advantages.

Deepening execution

The execution stack is already industry-leading; the next round of work makes it harder to match. Lower median latency on the routing path through co-located RPC infrastructure in additional regions. Smarter MEV mode selection that adapts to per-market conditions rather than per-trade size alone. New private execution paths with additional block builders to reduce single-relay concentration. The user-visible effect is the same: tighter fills, fewer surprises, more transparency.

Deepening discovery

Pulse, Discover, and Trader Scan are powerful surfaces but they reward expertise heavily — a new user needs hours of practice to extract their full value. The next round of work makes the discovery stack more legible without dumbing it down. Saved filter presets shared by the community, with provenance and performance attribution. Better explanations of why a given candidate scored the way it did. Onboarding flows that walk new users through their first real Pulse session rather than dropping them into the firehose.

Deepening portfolio

Cost-basis accounting and PnL are accurate today; what's missing is the longitudinal view that helps traders learn from their own history. Per-trade tagging with reasons, automatic clustering of trades into strategies, and post-hoc analysis that shows you which of your strategies actually generates positive expected value. The point is not to replace your trading journal; it's to make the trading journal something that maintains itself.

Expanding programmatically

The API exists today; the next round of work makes it a serious surface for non-developer traders. Strategy templates configurable without code — dollar-cost averaging, trailing-stop ladders, condition-triggered entries — with the same execution guarantees as manual trades. Webhook integrations with the tools traders already use. A library of community-contributed strategies with transparent performance histories.

What we will not do

We will not become a centralized exchange. We will not take custody of user funds under any rebranding of the same model. We will not introduce features that compromise the transparency commitments above. We will not expand into markets where we cannot deliver the same custody and execution quality we do today.

How to know if we're succeeding

Watch the platform's median fill latency, published quarterly. Watch the realized slippage distribution across the user base, also published quarterly. Watch the cadence and substance of the blog. Watch the team page; small teams that ship serious infrastructure are rare and worth tracking.

If those numbers go in the right direction, we're doing our job. If they don't, you should switch platforms and tell us why.