We all know how the cryptospace thrives on data and how critical its is to the industry. All sorts of solutions have been developed to help solve the any data problems developers face, especially when scaling their projects.
One solution has been able to beat many others by a long mile.
So much so, the heavyweights rely on it to bring out the best data-wise.
Avthar Sewrathan, lead technical product marketing manager at Timescale explores this and more.
Avthar Sewrathan, Lead Technical Product Marketing Manager, Timescale
What makes Timescale different from other time-series solutions?
At Timescale, we’re re-architecting PostgreSQL for modern applications. Timescale gives developers higher inserts, faster queries, millisecond response times via continuous aggregates and 90%+ storage cost savings. In turn, this allows developers to build more performant applications, and ship applications safer and faster than competing options via our enhanced cloud platform.
Timescale offers a more performant, more reliable and easier to use time-series solution than other cloud databases catering to time-series on the market today. Timescale has built in optimizations like hypertables and continuous aggregates that speed up ingest and query performance for relentless streams of time-series data, offering 350x faster queries than Amazon RDS and up to 7,100% faster than InfluxDB. Because Timescale is PostgreSQL, developers get the rock-solid reliability and stability of a battle-tested database, plus the ease of use of full SQL, the PostgreSQL tooling ecosystem, and hyperfunctions – Timescale’s custom SQL functions that make complex time-series queries easy to write.
Why choose PostgreSQL?
Developers love PostgreSQL, and for good reason. It’s battle-tested, having been around for over 30 years. It’s known and trusted, versatile, and has a really rich ecosystem of tools and connectors. PostgreSQL is today’s most advanced and most popular open-source relational database. We believe this as much today as we did five years ago, when we chose PostgreSQL as the foundation for Timescale because of its longevity, extensibility, and rock-solid architecture.
How does Time-Series as a data handling process integrate into the development of crypto and Web3 applications?
Time-series data captures how something changes over time. You use time-series data to analyze the past, monitor the present and predict the future. That kind of data is abundant in the crypto and Web3 world. The prime example is crypto asset trading data. The prices of these assets change every second and trades are happening every second. If you’re building an application that leverages that data, you’re going to want to surface insights about price changes, trading activity and other information to your users. So collecting time-series data enables you to 1) capture data at high fidelity and 2) create application experiences that leverage the time-based nature of that data.
Another example is blockchain transaction data. If you think about a blockchain as a data structure, it’s essentially a timestamped append-only log, and so everything on the blockchain is essentially time-based as well. That’s where the time series component comes in – you’re looking at activity over time and looking at how the network is being affected on a day-to-day and month-to-month basis.
The analogy I like to give is when you’re building these applications, you don’t just want to give your users a snapshot of the current state of the blockchain, but rather the whole movie of how it changed over time and how the network is being affected. So that time-based nature of time-series data is so critical to crypto and Web3 applications today. And that’s also why data infrastructure for time-series is critical as well.
Why use a database solution instead of a blockchain plus Oracles or other innovations that could scale?
Databases have a special place in the crypto and Web3 world. Part of the ethos around crypto is around decentralization and transacting in a trustless manner and blockchains are great for that. You don’t have to go through, or trust, any third party. But the way databases come into the picture is on the application layer.
When you’re building things like real-time monitoring, historical analytics, transaction history, insights, logging and other experiences that crypto users want, you’re going to want to use a database like Timescale. For these sorts of application experiences, using a blockchain and traversing the raw data on the blockchain would be too slow. For example, if I wanted to look at all my Bitcoin or Lightning transactions over time, it would take too long to aggregate that if I’ve transacted over years – I’d have to traverse a large percentage of the blockchain to get accurate data. So, databases serve that application layer rather than the transaction layer, where they can be used to store, aggregate, and analyze information that can then be used to create really rich user experiences.
For users, they can make transactions, trading and deployment decisions based on that data. So I believe databases are the backbone behind data rich application experiences that tell you stories and give you insight about the trustless transactions that take place on the blockchain layer.
Please, can you tell us about your cadre of customers?
Timescale works with thousands of customers – from startups to large enterprises – building innovative, data-centric applications. We power applications in markets spanning AI, crypto, IoT, energy, transport, financial services, and more.
Some notable names in the crypto space include: Messari, the data intelligence, analytics and research company, Hiro, the Bitcoin developer tools company, and Chainlink Labs, the web3 services platform.
Why work within the cloud?
The cloud has become the de facto way developers deploy and build applications today. In a world where you prioritize ease of use, specialization, and where the size of companies have shrunk and one developer can now do so much and have so much impact, cloud technologies go hand-in-hand with those shifts. It increases the leverage of an individual developer, where they don’t have to worry about how to provision a database, they don’t need to worry about back-up and restore, or creating database replicas. Everything is taken care of or available within one to two clicks. That’s the leverage developers get from the cloud, especially nowadays when you’re trying to keep your teams small and efficient. The cloud works well that way because it allows you to specialize and focus on your application rather than your database or other infrastructure that you’d rather have hosted in the cloud.
How do your API integrations enable the stability of enterprise solutions?
Timescale, as a cloud data platform for time-series, is built on PostgreSQL, which is battle-tested and tried and tested in many companies and production applications over the past thirty years. When you want to build enterprise solutions, you look for solutions that give you that stability, reliability and battle-testedness. Timescale takes PostgreSQL’s solid foundation and adds more performance and ease of use for time-series data, making it ideal for user-facing application experiences that enterprises are building. For example, one of our customers in the crypto space is Messari. Their APIs that their customers rely on are being powered by Timescale. Many views in their application, that customers are making trading decisions on or doing research to base their trading decisions on, that’s powered by Timescale. The performance and stability of Timescale comes from our query and storage optimizations, and our PostgreSQL foundation. Finally, as a Timescale user, you can use the whole ecosystem of PostgreSQL tools and connectors that’s been developed, that’s been used and extended by the PostgreSQL community. You’re not just locked in to what just Timescale can produce. You benefit from the whole community and that allows you to deploy your applications with confidence.
What are the benefits of deploying Time-Series data processes for financial applications, especially for the crypto space?
I think time series and finance have really been synonymous. During my undergrad when I was studying engineering, when I learned about time series data it was through the context of stocks and finance. And so crypto, being “internet money”, is a natural place where a lot of time series data is found. But in addition to financial use cases, there’s other examples of time series data that pop up in the crypto world. One is the blockchain transactions themselves, because the data contained within is time-stamped. It’s capturing the state of the network at any point in time and it allows you to look at how the network has changed over time. Also, consider the mining space and specifically Bitcoin, there are devices that are monitoring your mining output and looking at your hash rate and the health of your mining rigs etc. – all that monitoring data is all time series data. It’s all present in those three different layers – on the trading and price layer, on the actual blockchain transaction layer and then on the IoT layer of mining/device monitoring.
Those are the three examples of where time series data is in the crypto world and this is where you look for a database that can handle that time series data while also giving you that flexibility and reliability. That’s the niche that Timescale has carved out for itself. But I think that’s also why there’s an appetite for data infrastructure like Timescale because there’s so much of this kind of data in this space.
How critical are data processes to the growth and development of the crypto space?
As much as it’s taken the world by storm, the crypto space in general is still developing. In order to understand what’s working versus what’s not, companies rely on data. A recent example of this being talked about in the news is the scalability of Ethereum. That is only possible when you’re monitoring things like the transaction throughput of the network and the transaction speed of the network, all of which you need data for.
But in general, when we think about the growth and development of the crypto space, the market grows and evolves with every boom and bust cycle. The use cases that are actually useful live on and the others that might not have good product-market fit fade away and make room for new innovation. That is where data plays a part because with these new innovations and with every cycle, new kinds of data show up. A big winner from the last cycle was NFTs. NFTs came with their own masses of data, like real-time prices, floors, collection popularity, etc. And new companies popped up that used that data to help users – whether it’s marketplaces, trading platforms, wallets, etc. It’s not wrong to say that all the companies in the world need to leverage data to make the best products possible and crypto companies are no exception.
What do you think will be the effect of the recent crypto regulatory moves on data handling and storage processes?
As a crypto company, if you want to do business in a certain country, you have to abide by that country’s laws and regulations. Certain companies, like Gemini for example, have gone above and beyond to comply with that in the U.S. But in general, companies are going to have to place more emphasis on the security of their data and that’s where databases come in. Again, the data infrastructure that they choose needs to be able to keep up with evolving security and compliance measures – things like SOC2 and GDPR. The key for regulation is that it needs to provide a clear framework for companies to operate within, so that we can have the free markets that foster innovation. My wish is that the regulations are not overly restrictive and they don’t stifle innovation but promote it. The story is still playing out in terms of what the effects are actually going to be and it’s something everyone is keeping an eye on.
What do you think are the new use cases in the crypto space despite the recent challenges?
In the most recent cycle, the applications that gained a lot of adoption were NFTs and DAOs. NFTs were more of a success than many people thought they were going to be back in the mid 2010s and NFTs also brought a lot of mainstream adoption of crypto. Artists and musicians big and small leveraged them, in addition to communities that built token-gated applications and membership clubs. DAOs were another one where it was originally this theoretical idea around having a decentralized autonomous organization and then you started to see a bunch of DAOs popping up and collectively acting on certain things. One of the most famous examples of this was the ConstitutionDAO where people pooled their money together to try to buy the U.S. Constitution. The DAO didn’t end up winning the auction, but it was a really interesting idea that you could buy something collectively and keep ownership of it via a token. It’s interesting that we’re seeing these new use cases and those two specifically are seeing more real-world adoption than I think people thought previously.
Related: Cryptocurrency Investment Tips for Beginners
What is Timescale’s backstory?
Timescale founders Mike Freedman and Ajay Kulkarni had created a business called Iobeam which provided a full stack offering for IoT systems to connect to cloud data services. Essentially, the company provided a client SDK and server-side software for collecting client data. It was through this journey that the team found that none of the time-series database storage offerings suited their needs. The developers all loved the PostgreSQL database, but it couldn’t natively handle the volume and unique needs of time-series data very well. So the team set about creating a better way to store time-series data in a PostgreSQL database, and from that, Timescale was born.
Please, can you tell us about your pricing?
Timescale’s pricing is simple and transparent. We only charge for compute and storage resources. We allow for flexible scalability where you can scale your compute and storage resources independently. We make sure you’re getting the best price performance out of your database, where we have options to compress your data so it takes up less storage space and in turn, lowers your storage bill. We also offer data tiering where you can tier your data to a cheaper object store that will bring your bill down even lower. This is especially important in the crypto industry, where you don’t want to throw your data away. You want to keep it around as long as possible either to get the most complete picture of blockchain transactions or in terms of trades and prices, to have the longest time period of data to train models and make predictive decisions on. Timescale is well built for the use cases where you need to keep your data around for as long as possible.
How have Timescale deployments improved data handling efficiency?
Part of the value that Timescale brings is that it allows you to do advanced analytics, aggregations, and down sampling in the database layer. The advantage of this is that it’s more efficient and simpler than doing analytics, aggregations and other operations in the application code. I like to give a couple examples of this. One is data downsampling and aggregation. Say you collect data every second and and your users want to know the average transaction amount that took place in one hour or in one day. You need a way to roll up that data and Timescale has features that can do that for you – namely continuous aggregates. For example, for the data roll ups, you’d use continuous aggregates, which are basically automatically refreshing views where data is collected every second and it’ll automatically roll that up into a daily aggregate or a monthly aggregate or whatever is most interesting to the user. That’s one example where it’s much more efficient to do that in the database rather than in the application layer.
The other example relates to the finance and crypto world and that’s candlesticks. You can build candlesticks that tell you the open, high, low, close and volume (OHLCV) of a specific asset over a period of time directly in the database. Now that allows you to keep your candlestick as a data type in the database, and you don’t have to keep reconstructing that in the application layer. It makes it a lot easier to deal with because then you can serve that data in a form that is relevant to your users, rather than having to waste compute resources on the application side and slow things down.
How can developers integrate their solutions with Timescale’s data solutions?
The most common way is to use Timescale as your application database. For example, if you’re building a crypto intelligence tool or some sort of application that processes data and makes that available for users so that they can gain insights and make decisions from it, then Timescale is probably the best database to use to power that application. Meaning that the data that is used by that application is stored in Timescale and your user-facing views, dashboards, APIs can be served from Timescale. It provides a low latency experience with relevant data for your users. In terms of integration, whatever works with PostgreSQL works with Timescale so you can use any of the PostgreSQL libraries, drivers, connectors, tools to actually do the integration from your data pipeline to whatever you’re using from your frontend. It’s very easy to get started, we have tons of documentation on our website that developers can leverage when playing around with it.
What are your predictions for the crypto space as summer approaches?
Obviously 2022 was a very tough year for crypto with things like the FTX collapse, the Terra-LUNA crash, and several firms going bankrupt and having to shut down operations. My prediction/hope is that this doesn’t scare people away from the value that crypto can bring in terms of its useful applications, in terms of payments, NFTs, decentralized financial services and other things like that. I hope builders keep building applications that are useful. I think that will go a long way to help the market recover, as well as help end-users’ appetites for using crypto and crypto based applications to recover as well.
Do you have any other cloud-based data solutions/infrastructure in the near future?
Given the market conditions and the value of hierarchy of companies these days, everyone is looking at efficiency and price performance. At Timescale, we’re trying to build more things that align with those user preferences. For example, instead of over-provisioning resources and buying more than you need, we’re exploring various consumption based models where users only pay for what they use. We’re also thinking about other ways that we can give users better price performance and efficiency and look at where waste is occurring and how we can eliminate it in the data infrastructure layer. Those are the themes we’re seeing and we’ve started to build on. We plan to share more details in the coming weeks and months.
What are your expectations for Bitcoin Miami 2023?
I’m very excited for Bitcoin Miami. I’m happy that the community is getting together to look at the state of crypto today. This year there are a lot of panels and talks about the theme “What Happened”, for example the FTX collapse and what the aftermath is coming out of last year. My expectation is that it will be a very reflective conference. People usually don’t associate that with crypto, it’s a very fast-paced industry. But I do think it’ll be a conference where we take stock and look at how we can improve as a community going forward. I’m also looking to celebrate the spirit of innovation that’s alive in the crypto space. I’m excited to meet companies and software developers at those companies that are building useful things. That always energizes me and I think that’s the most important part of the event for me and for Timescale. Bitcoin Miami and the satellite conferences around it are a testament to the fact that yes, the industry has gone through a tough time, but the spirit of innovation is alive and well.
Related:Is it Too Late to Invest in Bitcoin? 6 Things You Need To Know First