*This text is taken from our video interview that you can find on our YouTube page.
The Big Whale: What led you to create RedStone?
Marcin Kazmierczak: My studies focused on quantitative methods and economics. At the end of 2017, I devoted my bachelor's thesis to blockchain, Bitcoin, Ethereum and the beginnings of smart contracts. In 2018, I joined a blockchain accelerator in Poland, aiming to explore new use cases for this technology. Although this project did not come to fruition, it fuelled my passion. After that, I held product manager positions at various companies before joining Google Cloud.
It was in 2020 that a friend suggested I start an oracle company in the crypto space. And so RedStone was born. Today, we're a 28-strong team, and we're proud to be the fastest-growing oracle on the market in 2024.
Despite your growth, Chainlink remains the ultra-dominant historical player in this segment, currently boasting over 40% of the market. How do you explain this dominance?
Oracles benefit from significant scale effects: the bigger they get, the faster their development accelerates.
Chainlink launched in 2018, at the height of the blockchain craze. Perfectly positioned, the project has capitalised on this opportunity. It became one of the catalysts of the "DeFi Summer" in 2020, enabling the rise of giants like Aave and Compound.
However, despite its revolutionary nature, Chainlink was not designed for the diversity of today's networks. At the time, we imagined a single dominant chain. Today, the multiplicity of chains requires a more scalable solution.
That's why we created RedStone with a modular architecture, capable of adapting quickly to new networks. We can integrate a new network in one to two weeks, compared with three to four months for Chainlink. This is a considerable advantage.
How do you explain this difference?
Chainlink's ultra-dominant position enables it to take a long-term approach with its customer base, mainly institutions in the traditional financial sector. They invest heavily in education, encouraging the adoption of crypto and decentralised finance (DeFi). This is crucial for the growth of the ecosystem.
As a small team of less than 30 people, we focus on specific problems rather than global issues as Chainlink does. Our responsiveness means we can quickly solve targeted problems for crypto protocols. This is one of our key growth drivers. For example, we have positioned ourselves as the fastest oracle to deploy on new chains. In the crypto industry, where speed is of the essence, an integration time of four months can be a serious handicap against the competition.
In addition, we offer both push and pull oracle models, the two most popular implementations today, whereas Chainlink only offers the push model. For a chain that wants both options, RedStone is the optimal choice. Finally, our team of engineers is constantly innovating, enabling us to adapt quickly to any new developments in the market.
Can you describe your "push" and "pull" models?
The "push" model aggregates data and transmits it to the blockchain at fixed intervals, depending on two conditions. Let's take the price of the US dollar as an example. Let's imagine that Chainlink updates this price every 24 hours. Even if the price fluctuates significantly during this period, the update is not immediate.
The second condition is the deviation threshold. This is the price deviation on external sources that triggers an update on the blockchain. For example, with an initial price of $1,000 and a threshold of 1%, the system only updates the price if it exceeds $1,011 or falls below $989.
Updating prices on multiple chains poses an additional challenge. Managing 20 chains means pushing updates to each one, which is costly and complex, especially for multiple assets.
That's why we designed the pull model. Instead of regularly transmitting data onto the blockchain, we only provide it on demand. Take the example of the Gearbox protocol: when a user initiates a transaction requiring a price feed, Gearbox asks RedStone for the current price. The user then incorporates this price feed into their transaction.
Imagine sending a letter to Gearbox to execute a transaction. Before it is sent to the blockchain, a small packet containing the price feed is attached. The user bears the additional gas costs for including this package, motivated by their desire to execute the transaction. This is proving to be more efficient, and in 2024 we are seeing increasing adoption of this pull model by protocols, due to its superior cost effectiveness and scalability.
Concretely, what important narratives have you had a significant impact on with your model?
A great example is Bitcoin staking, a hot topic in Q4 2024. We are collaborating with protocols such as Babylon and Lombard for liquid Bitcoin staking. RedStone has emerged as the oracle of choice for these projects, mainly because Chainlink couldn't meet their needs quickly enough. As a result, if you are looking to access liquid Bitcoin staking assets, RedStone is currently the most efficient solution.
According to DeFi Llama, the share of assets you secure has grown from just over 3% to almost 14% since January 2023. How do you explain such growth?
Since our launch on the mainnet in January 2023, we have operated for nearly two years without the slightest handling incident or pricing error. That's a record that not all our competitors can boast. Take Chainlink, for example: last December they mispriced one of DeFi's biggest assets, the Wrapped Staked ETH. I'm not claiming that we're infallible, mistakes can happen at any time, but we've always made security our top priority to minimise risk.
This reliability, combined with the strategy I mentioned earlier, has earned us an excellent reputation. In the world of DeFi, word of mouth is king. Landing our first customer was a challenge, but today our partners include leading protocols such as Morpho, Pendle , Venus, Sommelier, Frax, Lido, EtherFi , and many more.
What is the business model for an oracle company like RedStone? How do you generate revenue? Are you paid solely through contracts, or are there other methods?
In the "push" model, once the price stream is pushed on-chain, it becomes accessible to everyone. On public blockchains, it is virtually impossible to restrict access to this data. In theory, we could use Zero-knowledge Proofs (ZKP) technology, but this is costly and unusual in the way oracles work. So monetising on-chain data directly is proving complex.
We generally have three types of customer. Firstly, the blockchains themselves. They have an interest in supporting our infrastructure so that protocols built on their chain can access our price streams. Our customers include Zerocode, Storyprotocol, Monad, and other chains that want our presence in their ecosystems to support their developers.
Secondly, asset issuers, such as EtherFi, Ethena, or Lido. These protocols, wanting to use their assets in DeFi, contact us to create price feeds, for example for Wrapped ETH on a specific channel. We provide the service, sometimes under contract. Often they don't pay directly or pay a modest subscription fee, because their priority is growth, and we aim to cultivate good relationships with them.
In the "pull" model, an additional revenue stream is emerging. We could add a small transaction tax when the price feed is delivered. With each request, a small percentage would be added to the gas charges, reverting to the oracle. We haven't implemented this approach yet, as our current priority is growth. However, we are thinking about ways to future-proof our model as we grow.
Oracles seem crucial to industry adoption. This is a sensitive subject as it deals with on-chain price data. What is still missing from their development?
Smart contracts on Ethereum and other blockchains require oracles to integrate information from the outside world. Currently, oracles primarily provide market data - essentially financial asset prices, rather than information about elections or other aspects of the real world. This price stream is the core of their service today.
As the industry evolves, oracles will diversify their data types. The major challenge will be to create a viable business model based on these new offerings. For example, our infrastructure can provide GDP per capita for each country on-chain, but the willingness of customers to pay for this service remains uncertain.
I believe the biggest challenge for the DeFi industry and the blockchain ecosystem in general is to develop more sustainable business models for infrastructure providers like us. This is the key to expanding our services and supporting the industry's long-term growth.
One of the great strengths of the crypto ecosystem is its ability to innovate quickly. Can the same be said of oracles?
For oracles, reliability and security take precedence over rapid innovation. This priority naturally limits the speed and accessibility of the service, because of the inherent risks. Our customers demand absolute availability, a standard that exceeds even traditional web 2.0 business models. So caution is required in the delivery of the service.
I think the pull model could, in the future, solve many of these challenges, particularly in terms of scalability and gas costs. This model keeps data off-chain for as long as possible, only transmitting it to the target blockchain when a user expressly requests it.
The subject of oracles is also very important regarding the development of on-chain trading and in particular that of perpetual contracts on decentralised exchange platforms (DEX) . Can a DEX incorporating one of your oracles compete with the low latency of a centralised platform?
This is an excellent question that touches on the delicate balance between decentralisation and speed. In reality, the more decentralised or distributed a system is, the longer it can take to execute transactions or reach a consensus. Take the example of MegaEth, a Layer 2 that favours speed over decentralisation. They have designed an imposing physical machine with a high-performance sequencer, making it one of the fastest Layer 2s - simply because they have chosen not to focus on decentralisation.
Perpetual DEXs (PerpDex) face a similar dilemma. On Arbitrum, for example, the process goes like this: you first submit the trade, which is recorded on-chain, then the price is revealed and the trade executed in the next block. This delay acts as a built-in protection against front-running. However, there are always players looking to exploit these transactions for profit, making latency a crucial issue.
While I can't reveal all the details, we are currently working with cutting-edge PerpDex to develop a dedicated channel. Our aim is to reduce latency as much as possible while maintaining a high level of security. We're aiming for execution times of less than a second, or even half a second in some cases. That's almost as fast as Internet bandwidth and latency allow today, without compromising on security.
Can we know the names? dYdX , Hyperliquid perhaps?
I can't confirm anything now, but we hope to have something public by the end of the year, maybe early next year. The problem with perpetual DEXs is that they are extremely complex from an engineering point of view. You need incredibly low latency, and that's not just for an hour, it's 24/7, with maximum stability. It's not easy to create such implementations.
What are the current limitations for oracles?
There's an interesting concept we're working on that we call "Oracle Extractable Value" (OEV). It's similar to MEV (Maximal Extractable Value), and it's generated by oracle updates, particularly when a position on a decentralised application (dApp) is liquidated.
What characterises our solution is that most current solutions require the dApp to change its code or logic. Imagine platforms like Aave or Compound, for them to integrate OVC, having to re-audit and modify their operating logic. Our solution enables OVC to be activated without changing the logic of the dApp. We have abstracted the OVC implementation from the dApp, making it as simple as activating it.
Can you give a concrete example to illustrate how it works?
Imagine a loan position on Aave or another market that reaches its liquidation threshold. If a $1,000 position falls to $900, the liquidator can seize the asset, sell it and make a profit of $100. At the moment, this value is often captured by MEV bots that pre-empt liquidation. Using oracles, we can issue a price stream triggering liquidation with a signature that only authorises a specific party to liquidate. This ensures that the designated liquidator gets the value. We also organise a short auction between liquidators to determine who is prepared to pay the most for the opportunity - say $20 or $30. This auction takes place in milliseconds.
So, it's an auction between bots, not people?
Exactly. The bots are willing to bid for that specific price stream, and the payment is split between the lending protocol and us, the oracle provider. This generates a new revenue stream for everyone involved. If you have a sharp eye and know where to look on-chain, you might already see that this solution is up and running.
Does the CowSwap protocol work on a similar principle?
Yes, it's quite comparable. We are planning to implement two approaches: one where the protocol uses this configuration directly, and another offering more flexibility to the protocols. We're also in discussions with the CowSwap team about their MEV blocker, which follows a similar concept.
You're planning to deploy an AVS (Actively Validated Services) on Eigenlayer . What will it do for you?
The idea behind AVS is to add extra crypto-economic security for protocols that need it. For small projects, cost and speed of delivery take precedence over security. On the other hand, for large players such as Aave , Compound, or Venus, the potential cost of an attack is considerable. That's why RedStone will offer a variety of approaches for prize streams.
Currently, RedStone secures around $5 billion on blockchain. If that were to rise to $100 billion, the potential profit from an attack on our infrastructure would increase proportionately. This is where Eigenlayer's restaking comes in: it guarantees that any attack on our system would be economically unviable. For the most critical price flows, such as ETH/USD or Bitcoin/USD, this extra layer of security becomes essential.
Some flows will use the classic RedStone staking model. However, for more sensitive flows, particularly those linked to the USD, we will strengthen security with risk staking. In this model, risk-takers will protect the nodes supplying the price flows. If a fraudulent flow is transmitted, the node responsible will be 'slashed'. This approach ensures enhanced security, which is particularly crucial for high-value price flows.
How far along is the deployment process? Are you awaiting the launch of Eigenlayer's slashing mechanism?
At all. Our AVS is not yet on the mainnet because oracle slashing is crucial, and Eigenlayer has not yet deployed this functionality. However, we are also in talks with other restaking platforms, such as Symbiotic. We are keeping a close eye on developments.
Are you in contact with traditional financial players such as major banks or asset managers? Or is it still too early for these discussions?
We are currently in discussions with around fifteen institutional players. However, our main focus remains DeFi for the time being. Our team is relatively small - around 28 people, 70% of whom are engineers. So we work directly with players who have a clear vision of their needs.
We're growing fast and we're looking to the long term. Our ambition is to support major global players, such as JP Morgan, Fidelity and Goldman Sachs, in their transition to blockchain with concrete use cases. At the same time, we are preparing the launch of our token later this year. This will play a key role in restaking and other applications.
You have built up an impressive network, particularly with European players like Morpho and Angle. Is France an important market for you?
Yes, absolutely. Our biggest community is actually in France. They even organise local events in Paris. That's why we've created a dedicated Twitter account, RedStone France.
In your opinion, how soon will we see mass adoption of DeFi? And what role will oracles play in this process?
I think the mass adoption of DeFi will happen within two to three years. As for oracles, they will become similar to cloud services like Google Cloud or AWS: people will understand the concept, but won't necessarily interact with them directly.
For DeFi adoption, I think the inflection point will come when the big players start offering concrete solutions. For example, Securitize is doing great work with BlackRock, and Revolut recently announced plans for stablecoins. Once these specialist players start to demonstrate the value of their solutions, the others will follow. Imagine a service like Uber running on blockchain, where you pay in USDC - these are the kinds of use cases that could trigger mass adoption.
