Find Value with a Data Platform Approach

Application programming interfaces allow you to record raw data in forms and store it in databases, and automated procedures allow you to extract meaning from that data (APIs). The present process-centric attitude thinks that the value is in the automated processing, but the constraints and costs of relying on APIs have forced a rethinking of priorities and fresh thinking about where we can find data value. (data science in Malaysia)

Companies have learned the true cost of establishing and maintaining APIs as apps and their data structures have changed and grown over time.

The average cost of a business app is $175,000 to develop. According to him, creating a single API costs over $20,000 and maintaining it costs roughly 50% more per year. He noted, “So just implementing ten APIs will wind up costing more than the entire application.” According to him, the cost of the ecosystem required to maintain those APIs may be five to 10 times that amount. With some apps that have a lot of APIs and a lot of users, the cost of this “workaround” approach can be hundreds of dollars.

 Hackers are targeting APIs. (data science in Malaysia)

Security is another issue with APIs, in addition to rising expenses. According to a recent Akamai analysis, API assaults accounted for 75% of all hacking efforts against the financial services industry, posing a risk as well as higher security expenses.

When data is dispersed across numerous places and systems, as it is today, service applications are required to bring it all together and make sense of it, according to Platz. Even a small business today has upwards of 500 software and service applications, with many of these software service apps storing duplicated data.

Companies are turning to data lakes in an attempt to integrate their data sources: “An enormous cost, and they provide yet another attack surface.” In reality, he claims that data lakes and data warehouses have been the sites of some of the most serious data breaches in recent years. Because data security does not reside with the data, he claims that the process of moving data from one repository to another circumvents the data security embedded into the application tier. “It has no mechanism of defending itself if that we access a massive bank of data.” That’s a major issue.”

Data Management (data science in Malaysia)

Another basic fault with current It Management technology, according to him, is that there’s no way to tell where data came from, who placed it there. As more businesses rely on data to make vital choices, machines and humans alike must be able to verify that the information they are viewing is accurate. Data Quality is critical in domains such as healthcare.

A Novel Approach to Data

According to Dave McComb, author of Software Wasteland: A Novel, an application-centric approach focuses on a system’s function, which is generally described as a “business process.” How the Application-Centric Mindset is Getting in the Way of Our Businesses “This is a state of mind.” We are so engrossed in it that we are oblivious to it. This, we believe, is normal. We will be blind to the answer unless we see the problem.”

We’ve made it practically impossible for any genuine sharing to happen by letting each process and application to establish their own data structures, he claims. Platz concurs:

“People should consider using a ‘data first’ strategy.” Companies may either keep devising more sophisticated and expensive workarounds to make an application-centric strategy function in a data-centric world, or they can adopt a new data-centric approach.”

Data-Centric Architecture is a type of architecture that focuses on data.

The recognition that a data-centric architecture must take on a great deal of functionality that has traditionally been given over to applications, such as authentication, authorization, identity management, constraint management, query federation, and much more, according to McComb, is perhaps the most profound shift in Data Architecture. In a data-centric design, the architecture’s goal is to handle these functions once, at an enterprise level, rather than thousands of times and inconsistently per application.

We found Fluree in 2014 by Brian Platz and his partner Flip Filipowski. “There are many amazing technologies out there that have improved over time, but they also have a legacy that limits what they can do.” Platz and Filipowski chose to launch the company with a data-centric mindset so they could take a fresh approach to concerns like rising expenses and security vulnerabilities.

Fluree is an immutable, time-ordered blockchain database powered by an RDF-graph database engine, according to the company’s website. We cryptographically signed and connected each block which that is an atomic update to the previous block in the chain to avoid manipulation.  Users can scale components independently rather than having to add new servers in order to expand.

Source: data science course malaysia , data science in malaysia