[ad_1]
Data has emerged as one of the world’s greatest resources, underpinning everything from video-promoting engines and digital banking to the burgeoning AI revolution. But in a world where data is increasingly distributed everywhere, from databases to data warehouses to data lakes and beyond, combining it all into a compatible format for use in real-time situations is a daunting task.
For applications that do not need contextual, real-time, real-time access to data, they can easily aggregate and batch the data at set intervals. This so-called “batch data processing” can be useful for processing data such as monthly sales. But often, company will do It needs real-time access as it happens, and this can be key for customer support software, for example, that relies on up-to-date information about each sale. Elsewhere, ride-hail apps need to process all sorts of data points to connect a rider to a driver — not something that can wait a few days. Such scenarios require what is known as “streaming data processing” where data is collected and combined for real-time access – something that is quite complex to set up.
And that’s exactly what Dozer aims to solve, providing fast, read-only APIs directly from any source behind a plug-and-play data infrastructure.
Dozer in manual work Vivek Gudapuri And Matteo Pelati, who founded the company a year ago in Singapore. The duo has built a team of 10 across Asia and Eastern Europe as they prepare to expand from an existing product (ie non-open source) to a fully monetized product.
Dozer has been testing the product with a few unknown design partners, and today it’s coming out of the closet for any developer to access. The company also announced that it has raised $3 million in seed funding from Sequoia Capital India, Google Gradient Ventures, Surge and Janwari Capital.
Dozer co-founders Matteo Pelati and Vivek Gudapuri Image creditsDozer
It has been distributed.
There are countless tools designed to transform, integrate, and consume distributed data, including streaming database and ETL (extract, transform, load) tools such as Apache Flink, Airbyte, and Fivetran; caching of temporary data storage layers such as Redis; And fast APIs powered by the likes of Hasura or Supabase to transfer data between systems.
Dozer, on the other hand, works in all these different categories, taking what it thinks are the best and avoiding the friction of building the infrastructure and pipelines that support real-time data applications.
Users plug Dozer into their existing data stack, which can include databases, data warehouses, and data lakes, and Dozer takes care of data extraction, caching, and indexing in real-time and processes it through low-latency APIs. So while something like Airbyte or Fivetran helps put data into a data warehouse, Dozer focuses on the other side — “making that data accessible in a very efficient way,” Gudapuri explained to TechCrush.
Guadapuri says that Dozer “takes a conceptual approach”, which solves very specific problems and is not additive. For example, the current streaming database solves many problems beyond what Dozer offers, all in one product by providing real-time data updates and APIs.
“We solve the right amount of problems in each of these categories to provide developers with a faster build experience as well as ready performance,” said Godapuri. Developers (currently) have to integrate multiple tools to achieve the same thing.
For example, a current streaming database probably tries to provide the entire database experience to the user, complete with query engine, data search, OLAP (online analytical processing), and so on. Dozer intentionally doesn’t offer these things, instead focusing on what Pelati calls “pre-computational views” using SQL, Python, and JavaScript, all available via low-latency gRPC and REST APIs.
And that’s why, Pelati said, Dozer can promise better data query latency.
“Because of these design choices, Dozer delivers the extremely high request latency necessary for customer-facing applications,” Pelati said. “A single developer can spin up entire data applications in minutes, which would normally take months. A team does not have to build and maintain multiple integrations, saving time and money.
Because of open source (not generally)
While Dozer is being touted as “open source,” a look at its license on GitHub shows it uses the Elastic License 2.0 (ELv2), the same license search firm Elastic adopted as part of its transition two years ago. Far away From true open source. In fact, the Elastic License is not recognized as open source, as it prevents third parties from taking the software and offering it themselves as an integrated or managed service.
More precisely, ELv2 can be called an “available source” license, which means that it offers many advantages of a more voluntary open source license like MIT, including codebase transparency, the ability to extend Dozer’s capabilities or fine-tune features and fix bugs. That alone could be enough to win the hearts and minds of businesses of all sizes, unless it’s AWS or another cloud giant looking to monetize directly on top of the dozer.
However, the company said it plans to switch to a dual license “soon,” with everything in the Core Dozer project being licensed by MIT except for “one core module.” What’s more, the company emphasizes that all of its client libraries, including Python, rect, and JavaScript, are MIT-licensed.
It’s worth noting that some companies, including Netflix, which built Bulldozer several years ago, have developed their own internal tools to solve the same problem that Dozer is facing. Notably, one of the main inventors behind the bulldozer, Ioannis Papapanagiotou, now works as a dozer consultant.
It’s still early days for Dozer, but with $3 million in the bank and a host of high-profile backers, the company is well-positioned for a commercial push, including the introduction of a multi-suite SaaS version. Additional features. Gudapuri said they expect this to go live in the coming months.
“The hosted service takes care of auto-scaling, rapid deployment, security, compliance, rate limiting and some additional features,” Gudapuri said.
[ad_2]
Source link
