How Omniston Scaled from Liquidity Layer to Execution Engine in Production

by

From Liquidity Layer to Execution Engine: How Omniston Scaled in Production

Creating a decentralized exchange (DEX) application isn’t too difficult. However, successfully operating it in a live, fast-moving market – with automated trading programs, arbitrage traders, and rapidly changing funds – is a major challenge. BeInCrypto interviewed Andrey Fedorov, CMO & CBDO at STON.fi Dev, at Consensus Hong Kong to learn about the real-world experience of launching and running such a platform.

STON.fi started as a way to swap cryptocurrencies directly on the TON Blockchain, using what’s called an automated market maker. Later, they created Omniston to solve a problem: with several different exchanges (DEXs) on TON, users had to check prices on each one individually. Omniston aimed to simplify this by bringing all the available funds together in one place, making it easier to find the best prices.

Aggregation worked. But scale exposed new constraints.

Three Lessons From Production

Fedorov readily admits they faced challenges as things grew. Initially, it was simple – they only had one type of access key and very few users. However, demand quickly increased, leading to a surge in activity.

We quickly learned about scaling – both the user-facing side and the underlying systems struggled when demand unexpectedly surged. Another challenge was more nuanced: trades that went through multiple tokens worked well in our tests, but caused problems when used with real-world traffic. As Fedorov explains, while these multi-step trades *should* work perfectly, in reality, several things can go wrong at the same time. Transactions happen simultaneously, available funds in different trading pools change quickly, and multiple exchanges update their information all at once. This means the first part of a trade might succeed, while the second part fails.

The third key takeaway was the inherent complexity of the system. The original plan envisioned a straightforward exchange: users would trade, and liquidity providers would offer funds. However, the real world introduced additional players like arbitrageurs and bots, along with unexpected interaction patterns. As one expert put it, “You can’t predict everything upfront. It’s better to launch, observe what happens, and address any issues as they arise.”

STON.fi currently handles 80-90% of all trading activity on the TON blockchain, making it the clear leader in swap volume. However, the developers are planning to add cross-chain swaps, which are expected to significantly change the landscape. While the core principles will remain, they anticipate facing new obstacles as a result.

Why Aggregation Wasn’t Enough

Initially, Omniston planned to link all the liquidity pools on TON DEX to discover the most efficient trading paths. However, simply combining existing public liquidity has its limits. If there’s no liquidity available for a specific trading pair, even the smartest routing won’t be effective.

Sometimes, there simply aren’t enough funds available in a trading pool, as Fedorov explains. This means if someone wants to exchange tokens in that pool, they might not get a fair price because of the lack of available funds.

The solution was escrow swaps, which allow Omniston to access hidden liquidity from experienced traders, called “resolvers,” alongside traditional public pools. This means Omniston now checks both public and private sources and chooses the option that gives users the best possible trade.

This isn’t a simple fix-all solution; we need both elements working together. Using them in combination creates the best possible result.

Tokenized Equities as a Stress Test

As a researcher, I’ve been observing how the escrow model really shone when STON.fi started working with xStocks. These are essentially digital versions of US stocks, created by Backed Finance, and built on the TON blockchain as jettons. However, they don’t act quite like typical crypto tokens, and those differences are important when it comes to how transactions are processed.

A bigger hurdle was ensuring enough liquidity. Unlike popular cryptocurrencies, xStocks don’t have large, readily available automated market maker (AMM) pools. While AMM support exists, we also created escrow swaps to give users access to more liquidity. Currently, the majority of xStock trades happen through this escrow system.

As an investor, what I really appreciate is that the team is aiming for a seamless experience. They want swapping crypto to feel just as easy as using any other platform – no complicated steps or technical jargon. They’re handling all the complex stuff behind the scenes so *I* don’t have to worry about it, and that’s a huge plus.

The Self-Custody Trade-off

Fedorov is direct about the constraints of remaining fully non-custodial. 

We often encounter potential solutions that are already popular and have many users. While adding these could quickly help our business grow, a lot of them rely on a central authority. When we present these options to our engineers, they explain that these solutions aren’t compatible with how we operate. STON.fi is different – it gives users full control of their assets, keeping them safely in their own wallets, and uses secure smart contracts to handle all transactions.

Connecting to traditional systems is usually quick and easy, often just requiring a simple API link. However, integrating with decentralized finance (DeFi) is different. It demands secure, code-based solutions where users always maintain control of their funds. As one developer put it, “We could grow more quickly if we took custody of user funds, but that would mean we weren’t truly building DeFi – we’d just be another financial technology company.”

As a crypto investor, I’ve realized it’s not just about the tech itself, but also about understanding how things work. This can be tricky to explain to people new to crypto. The biggest shift with self-custody is that *you* are in control of your funds, which is great, but it also means you’re fully responsible. A lot of newcomers don’t realize this – they think if they lose their seed phrase, someone will magically fix it for them. It’s different than a bank or an exchange; we can’t help if you lose that phrase, because we never had access to it in the first place, but people still expect us to! It’s a communication challenge, for sure.

Traditional systems offer ways to help users who lose access to their accounts, like password resets and customer support with override access. Decentralized Finance (DeFi) relies on security through eliminating those types of ‘backdoors.’ This means the features that keep users safe also prevent us from stepping in to help when problems arise.

For STON.fi, this means focusing more on helping new users get started, providing better educational resources, and making the platform easier to use – all while still giving users complete control of their assets.

This is an investment for the future. While educating people about it can be challenging right now, eventually users will appreciate the benefits of truly owning their data and assets. That’s especially true with Web3, where ownership is central to the idea.

Distribution First, Then Depth

Fedorov sees TON as more than just a blockchain; it’s a way to reach a large audience thanks to its connection with Telegram. Platforms like STON.fi and Omniston work with various Telegram wallets, apps, games, and bots, creating many opportunities for users to trade directly within those services. This benefits everyone involved: developers can easily add trading features to their apps, and TON gains a built-in distribution network.

Our next step is to combine liquidity from different blockchain networks, beginning with Tron and then extending to other EVM-compatible chains. This will allow us to pull together funds from across multiple ecosystems, not just from different exchanges within a single blockchain.

Fedorov explains their plan is to simplify things for non-technical users, expand access by working with all popular apps, and combine resources from various blockchains, instead of being limited to just one. He says the next step is to grow and expand this system.

Read More

2026-02-17 16:07