Charting a Course to Enabling AI-Powered DeFi: Our Roadmap for Q1/Q2 2024
EmpyrealSDK is on a mission to radically simplifying user journeys and removing friction through our groundbreaking tech.
Our ambition is to become the new standard for intelligent infrastructure by building infra that enables every DeFi player — protocols, dApp developers, power users — to build and deploy AI-powered onchain services.
EmpyrealSDK is realizing this ambition by bringing autonomous agents to life with scalable and flexible automation, real-time data streaming, and account abstraction that facilitates a trust-minimized environment.
Let’s look at how EmpyrealSDK has ideal product-market fit with DeFi protocols and dApps, the market opportunity, and how our roadmap for Q1/Q2 2024 is getting us closer to becoming the infra powerhouse behind premier DeFi products.
Solving persistent challenges for DeFi
EmpyrealSDK’s focus on real-time data streaming, intuitive autonomous agents, account abstraction, and snap-to-grid modular design give us ideal product-market fit for DeFi protocols and dApps.
The challenge and the opportunity for DeFi is creating real-time, block-by-block automations based on streaming real-time data, and doing so in a delegated, trust-minimized environment.
Access to accurate real-time data is essential to be able to make decisions around yield, price, and other variables — for both users and protocols.
Data needs to be processed on the fly by trained models to determine the best action to take in pursuit of yield-bearing and other opportunities.
Onchain actions have to be executed in a timely manner to take advantage of these time-sensitive opportunities.
And those onchain actions need to be executed in a delegated, trust-minimized environment to preserve the safety of funds and avoid compromising wallet or protocol security.
That’s exactly why we are building EmpyrealSDK, and our solution has many applicable use cases across the DeFi space today:
Yield optimization
AI trading
Analytics for protocols and users
MEV optimization
any use case that requires data and automation with trust-minimized delegation
Market opportunity for AI-powered DeFi
In last cycle’s peak, DeFi TVL reached $108bn on Ethereum (Nov 2021). Today there is $43bn TVL on Ethereum DeFi alone (Feb 2024). We are expecting the current cycle’s peak to be multiples higher than 2021 because of institutional adoption, so estimate that Ethereum alone could see DeFi TVL between $250bn and $350bn by 2025.
By offering their users block-by-block automations based on real-time data streaming in a trust-minimized environment, DeFi protocols would be adding another source of revenue and creating a positive engagement flywheel, allowing users to select a premade trading or portfolio rebalancing strategy or to design a custom one of their own.
By building dApps and products that take advantage of real-time data streaming in a trust-minimized environment to allow users to create their own block-by-block automations, developers would be adding another source of revenue to their trading, portfolio management, or yield optimization tools.
In all the examples above, EmpyrealSDK is the one stop shop to provide these services out of the box, with minimal hassle and easy snap-to-grid modular integration into any protocol or dApp. We’re building all of these components, fully integrated, from the ground up — giving us a significant advantage over others who are working on them piecemeal.
How much of that $250bn and $350bn TVL will go in search of the best possible yield on a daily or weekly basis, want to execute a grid trading strategy on large assets, seek to autocompound yield from one protocol and deposit it into another, and countless other actions rooted in real-time data and timely on-chain execution?
We see a tremendous opportunity for EmpyrealSDK here, and we’re just at the very beginning of our journey to conquering DeFi and web3. Our business model is transparently outlined in our white paper: we collect fair fees on every transaction facilitated by EmpyrealSDK, and we share 50% of those fees with $EMP stakers.
Validating our solution & forming partnerships
Our conversations with major protocols have only confirmed that we’re working in the right direction, and that we’re solving problems that nobody else has been able to solve. We can’t wait to reveal some of the projects that we’re going to be partnering with in the coming weeks and months.
We’re continuing to execute against our two-pronged partnership strategy: top-down to tier 1 protocols that already have tens of thousands of users and hundreds of millions in TVL, and bottom-up with eager developers looking to build new products incorporating EmpyrealSDK.
Roadmap Q1/Q2 2024
Our roadmap will take us closer to realizing our vision of becoming the standard for intelligent infrastructure by building infra that enables every DeFi player to build and deploy AI-powered onchain services:
AI Model Training and Data Feeds
Enhancement pipelines for our library of data inputs used to train AI models
Data feeds for 2 soon-to-be-disclosed protocols
LRT/LSD data feeds
Indexed data lake of raw Ethereum transaction & event data
Where we’re at
We’ve built indexing infrastructure that allows for serving complex data on transfers and swaps to our downstream clients. We have built this as a generalizable framework and are working on scaling this infrastructure to facilitate handling more data, faster.
We currently have capabilities for UniswapV2, UniswapV3, SushiSwap, and many other smaller DEXs on Ethereum and Arbitrum, perps from GMX and Orderly, and price data from CoinGecko, Binance, and ChainLink, and we will be working to improve the process for adding new data feeds seamless to our indexing and processing pipelines so we can support diverse, multichain requirements.
What’s next
We’re continuing to add more events to our indexing and processing pipelines so AI-powered strategies will have more information at their disposal because access to accurate real-time data is essential to be able to make decisions around yield, price, and other variables — for both users and protocols. And that data needs to be processed on the fly by trained models to determine the best action to take in pursuit of yield-bearing and other opportunities.
Architecture for Automation and Privacy
Pipeline migration to cloud native infrastructure for scaling up to millions of simultaneous agents processing Terabytes of data
Trustless deployment of proprietary trading strategies
Account abstraction to facilitate trustless delegation of transaction authority to autonomous agents
Custom selection of data feeds to power trading strategies
Backtesting of trading strategies against historical data
Where we’re at
We currently have infrastructure to manage data for our customers on metrics like swaps, token prices, transfers, and some additional customized metrics. We’re migrating our pipelines to a cloud native infrastructure where we can scale to thousands of data processing pipelines for real-time analytics.
Our infrastructure is being built for a library of data inputs used to train AI models using industry best practices. This will allow for more rapid backtesting, feature generation, and mobilizing these features to train diverse models.
We have an architecture to support individualized trading strategies and onchain automation, which we’re now strengthening to enable deployment of millions of simultaneous autonomous agents.
What’s next
Building these items is what will drive our monitors, trading strategies, and metrics for our library of data inputs used to train AI models. By ingesting data real-time in our distributed architecture, we are able to ingest a large amount of data, process it in real-time, and feed this into our subscriptions that are used by our real-time autonomous agents.
We’re scaling our infrastructure to process more data in less time. As we process more datasets and begin to use them to trigger our onchain actions, we will be able to scale effortlessly to handling advanced analytics on real-time data.
And we’re building customizable strategies ingesting information from our diverse datasets for trading strategies, Maximal Extractable Value (MEV), and monitors for onchain events that can trigger processing pipelines to result in timely onchain actions for users of EmpyrealSDK.
—
2024 is going to be a very big year for Empyreal. The foundations we laid are enabling us to focus on scaling up — and we’re only just getting started.
Watch as we become embedded in the very fabric of web3, enabling every DeFi player — protocols, dApp developers, power users — to build and deploy AI-powered onchain services.
And through it all, Empyreal is unlocking value — one protocol, project and product at a time — and capturing it through our token.
Stay up-to-date with Empyreal: