Eco Built the Wallet. Then It Found the Real Business Hiding Underneath

Eco learned the expensive way that a clean wallet cannot make a messy financial system disappear. In 2021, the San Francisco company looked like a bright new consumer-finance entrant: one app for saving and spending, yields of 2.5 to 5 percent, 5 percent cashback at familiar merchants and no fees. More than 180,000 people joined the waitlist. Average deposits, Eco said then, reached $7,700. Venture firms supplied $86 million in two rounds that year, on top of an earlier $8.5 million.

That version of Eco had an easy sentence and a hard backend. It wanted digital money to feel useful every day. But as the team moved from the Eco App to Beam - later renamed Bend - a self-custody wallet billed as a global Venmo, it kept meeting the same obstacle. Stablecoins might all look like dollars to a person, yet they sit on different blockchains, use different bridges, require different gas tokens and draw from separate pools of liquidity. A transfer that begins with a dollar can end with a tutorial.

So Eco changed layers. The company now sells the machinery it wished existed while building the app: a stablecoin execution network that takes a desired outcome, finds a route, coordinates a solver, verifies the result and settles the money. The wallet did not vanish from the history. It became the field research.

The first thing to fail was coordination

Blockchains can process transactions. Bridges can pass assets or messages. Decentralized exchanges can swap one token for another. The failure arrives when all of those jobs must line up for a single payment. Liquidity moves. Quotes expire. A destination call can fail after funds have already left. Someone must front capital, someone must prove completion and someone must reconcile the result.

Eco's diagnosis is blunt: cross-chain payments fail first at the coordination layer, not the execution layer. Its flagship product, Routes, replaces a list of prescribed steps with an “intent.” The user or application states the outcome - for example, pay USDT on Solana using USDC held on Ethereum, within a slippage limit. Independent solvers compete to deliver that outcome. Eco's contracts publish the intent, escrow the reward, record fulfillment and release payment after proof.

The small but important trick is isolation. Each intent can use a dedicated deterministic vault, with an address knowable before deployment. Funds for one outcome do not dissolve into an anonymous pooled balance. That one-to-one mapping matters for a merchant payment or an invoice, where “the money generally arrived somewhere” is not acceptable bookkeeping.

“The industry has made real progress on scaling blockchains. It has made far less progress in making them operate together as a reliable financial system.”Eco, April 2026

Not a better bridge. A chooser of bridges

Eco's distinction is orchestration. Circle's CCTP, LayerZero, Hyperlane and other systems are rails. Eco sits above them, selecting a path and stitching execution into one interface. That makes the company a partner to some infrastructure providers and a competitor to aggregators such as LI.FI, Across, Socket and deBridge. The boundaries are deliberately untidy: LI.FI and its consumer interface Jumper also distribute Eco Routes.

Specialization is the wager. A broad route aggregator may support more long-tail assets. Eco tunes its system for stablecoins, where quote validity, predictable amounts and reconciliation matter more than showing every token ever minted. Its current pitch is aimed at wallets, protocols, issuers, treasury teams, fintechs and institutional money movers. In 2026 it added Verified Liquidity, a route for regulated institutions that need approved counterparties and policy controls.

Routes is the spine. The rest makes money programmable

The product line now resembles an operating system. Routes handles transfers and swaps. Permit3 extends token permissions across chains, so a user can authorize a bounded action without repeating the usual approval dance everywhere. Sauce is a workflow engine for conditional, multi-step operations. Programmable Addresses can trigger a swap, bridge, deposit or split when funds arrive. Accounts, advertised as coming soon, is meant to present one stablecoin balance across connected chains.

Then there is Crowd Liquidity, Eco's attempt to turn idle stablecoins into a shared source of capital for intent solvers. Instead of every solver maintaining money on every chain, capital holders can supply liquidity to where order flow needs it. Eco compares the design to an onchain order book. The economic promise is deeper routes and a new yield source; the caveat is the familiar one in DeFi - smart-contract, market and liquidity risk do not disappear because the interface is tidy.

The company does not publish a simple price card. Routes is permissionless to integrate, and solver competition is supposed to produce pricing rather than a single fixed toll. Enterprise APIs, routing and institutional controls are the likely commercial center, but Eco does not disclose a revenue mix. Its public employment footprint is also much smaller than some data vendors suggest: LinkedIn lists 11 to 50 employees, and Eco reported a team of 45 in 2023.

The network grows by borrowing distribution

Infrastructure does not win merely because its contracts are clever. It needs to appear where people already move money. Eco's expansion since 2025 reads like a map of borrowed distribution: LI.FI and Jumper put Routes in front of bridge users; Caldera included Eco in an aggregator spanning more than 40 rollups; Para paired its wallet permissions with Permit3. Everclear became a primary rebalancing partner, doing the unglamorous work of returning capital to the chains where the next orders need it.

The chain integrations have followed demand rather than a single technological tribe. Eco connected World Chain, Polygon, Celo and Sonic, then added Plasma for Tether-heavy flows and Solana for its large stablecoin economy. Hyperliquid, BNB Chain, TRON and Robinhood Chain followed in 2026. The point is not another row of ecosystem logos. Each addition makes a single Eco integration more useful to the apps already connected elsewhere.

That creates a familiar network-effect loop. More destinations attract more applications and order flow. More order flow gives solvers a reason to commit capital. More solver capital can improve price and speed, which makes the integration more attractive. But the loop is not automatic. A new chain without meaningful stablecoin demand adds maintenance before it adds liquidity. Eco has to choose expansions carefully enough that breadth does not become a collection of thin, expensive routes.

The company's open-source posture helps on the supply side. Routes implements the ERC-7683 cross-chain order standard and supports multiple proof systems. Developers can inspect the contracts, run solvers and build against documented APIs rather than accepting a black box. Openness does not remove contract risk, but it makes Eco easier to evaluate and harder to dismiss as a private switchboard.

What changed Eco's mind

The pivot was less a fashionable swerve than a sequence. The Eco App proved people wanted a friendlier dollar product. Beam/Bend proved account abstraction could hide seed phrases, gas and wallet addresses. But neither could cheaply solve fragmented liquidity underneath. Rebuilding route selection and settlement inside every app looked wasteful. The repeated constraint became the product.

Ryne Saxe, a co-founder who returned to the CEO role after serving as COO, now leads the infrastructure chapter. Andy Bromberg, who led Eco and Bend through the consumer years, describes himself as Eco's chairman. The corporate through-line survives their role changes: bring dollars onchain, then remove the crypto homework that keeps those dollars from behaving like money.

When the model is too much machinery

Eco is not the automatic answer to every crypto payment. A single-chain app with one stablecoin and deep local liquidity may be better off calling its native contracts directly. A trading product centered on obscure assets may need a broader aggregator. A company that cannot accept smart-contract or solver risk still needs conventional custodial rails, even if they are slower. And an unsupported chain remains unsupported, however elegant the abstraction above it.

There is also a market condition hidden inside the architecture: competition only works when enough solvers and enough capital want the order flow. Thin routes can still produce weak quotes. Crowd Liquidity is meant to improve that condition, but it must attract supply without paying incentives that overwhelm the value of the transactions it supports.

Still, Eco's journey offers a useful inversion. The company began with the thing everyone could see - a wallet, rewards, a friendly consumer promise - and ended up selling the part nobody wanted to think about. Its achievement will not be making bridges exciting. It will be making them forgettable.