Daily Tech Digest - June 30, 2021

DigitalOcean aligns with MongoDB for managed database service

There is, of course, no shortage of DBaaS options these days. DigitalOcean is betting that its Managed MongoDB service will not only extend the appeal of its cloud service to developers, but also to SMBs that are looking for less costly alternatives to the three major cloud service providers, Cooks said. MongoDB already has a strong focus on developers who prefer to download an open source database to build their applications. In addition to not having to pay an upfront licensing fees, in many cases developers don’t need permission from a centralized IT function to download a database. However, once that application is deployed in a production environment, some person or entity will have to manage the database. That creates the need for the DBaaS platform from MongoDB that DigitalOcean is now reselling as an OEM partner, said Alan Chhabra, senior vice president for worldwide partners at MongoDB. The DigitalOcean Managed MongoDB service is an extension of an existing relationship between the two companies that takes managed database services to the next logical level, Chhabra asserted. “We have a long-standing relationship,” he said.

Digital transformation at SKF through data driven manufacturing approach using Azure Arc enabled SQL

As SKF looked for a solution that supported their data-driven manufacturing vision for the Factories of the Future, they wanted a solution that was able to support distributed innovation and development, high availability, scalability and ease of deployment. They wanted each of their factories to be able to collect, process, analyze data to make real-time decisions autonomously while being managed centrally. At the same time, they had constraints of data latency, data resiliency and data sovereignty for critical production systems that could not be compromised. The drivers behind adopting a hybrid cloud model came from factories having to meet customer performance requirements, many of which depend on ability to analyze and synthesize the data. Recently, the Data Analytics paradigms have shifted from Big Data Analysis in the cloud to more Data-Driven Manufacturing at the machine, production line and factory edge. Adopting cloud native operating models but in such capacity where they can execute workloads physically on-premises at their factories turned out to be the right choice for SKF.

A new dawn for enterprise automation – from long-term strategy to an operational imperative

To drive sustainable change, organisations need to take a large-scale, end-to-end strategic approach to implementing enterprise automation solutions. On one level, this is a vital step to avoid any future architecture problems. Businesses need to spend time assessing their technology needs and scoping out how technology can deliver value to their organisation. Take, for example, low code options like Drag and Drop tools. This in vogue technology is viewed by companies as an attractive, low-cost option to create intuitive interfaces for internal apps that gather employee data – as part of a broad automation architecture. The issue is lots of firms rush the process, failing to account for functionality problems that regularly occur when integrating into existing, often disparate systems. It is here where strategic planning comes into its own, ensuring firms take the time to get the UX to the high standard required, as well as identify how to deploy analytics or automation orchestration solutions to bridge these gaps, and successfully deliver automation. With this strategic mindset, there is a huge opportunity for businesses to use this thriving market for automation to empower more innovation from within the enterprise.

The Rise Of NFT Into An Emerging Digital Asset Class

The nature of NFTs being unique, irreplaceable, immutable, and non-fungible makes them an attractive asset for investors and creators alike. NFTs have empowered creators to monetize and value their digital content, be it music, videos, memes, or art on decentralized marketplaces, without having to go through the hassles that a modern-day creator typically goes through. NFTs, at their core, are digital assets representing real-world objects. ... NFTs solve the age-old problems that creators like you and I have always faced when protecting our intellectual property from being reproduced or distributed across the internet. The most popular standard for NFTs today are ERC-721 and ERC-1155. ERC-721 has been used in a majority of early NFTs until ERC-1155 was introduced. With that said, these token standards have laid the foundation for assets that are programmable and modifiable; therefore, setting the cornerstone for digital ownership leading to all sorts of revolutionary possibilities. The NFT ecosystem has found its way into various industries as more people join hands and dive deeper into its novel possibilities. 

Three Principles for Selecting Machine Learning Platforms

Of the challenges this company faced from its previous data management system, the most complex and risky was in data security and governance. The teams managing data access were Database Admins, familiar with table-based access. But the data scientists needed to export datasets from these governed tables to get data into modern ML tools. The security concerns and ambiguity from this disconnect resulted in months of delays whenever data scientists needed access to new data sources. These pain points led them towards selecting a more unified platform that allowed DS & ML tools to access data under the same governance model used by data engineers and database admins. Data scientists were able to load large datasets into Pandas and PySpark dataframes easily, and database admins could restrict data access based on user identity and prevent data exfiltration. ... A data platform must simplify collaboration between data engineering and DS & ML teams, beyond the mechanics of data access discussed in the previous section. Common barriers are caused by these two groups using disconnected platforms for compute and deployment, data processing and governance.

Introduction To AutoInt: Automatic Integration For Fast Neural Volume Rendering

AutoInt, also known as Automatic integration, is a modern image rendering library used for high volume rendering using deep neural networks. It is used to learn closed-form solutions to an image volume rendering equation, an integral equation that accumulates transmittance and emittance along rays to render an image. While conventional neural renderers require hundreds of samples along each ray to evaluate such integrals and require hundreds of costly forward passes through a network, AutoInt allows evaluating these integrals with far fewer forward passes. For training, it first instantiates the computational graph corresponding to the derivative of the coordinate-based network. The graph is then fitted to the signal to integrate. After optimization, it reassembles the graph to obtain a network that represents the antiderivative. Using the fundamental theorem of calculus enables the calculation of any definite integral in two evaluations of the network. By applying such an approach to neural image rendering, the tradeoff between rendering speed and image quality is improved on a greater scale, in turn improving render times by greater than 10× with a tradeoff of slightly reduced image quality.

How Google is Using Artificial Intelligence?

In the old times, we were much dependent on the paper map or the suggestions of people well-versed with the tracks of our destinations. But with that, the problem was we never reached on time to our spots. Now, you need not seek such suggestions from the people or a paper Map as Google Maps has solved the related difficulties. With territories and over 220 countries like Delhi, the United States, Pakistan, Australia, etcetera one can affordably reach the places already decided. You may curiously ask about the technology embedded and the answer for this is Artificial Intelligence. The main concept is global localization which is relying on AI. This helps Google Maps understand your current or futuristic orientation. Later, it lets the application precisely spot your longitudinal and latitudinal extent and as you or your vehicle proceed further, Google Maps starts localizing hundreds of trillions of street views. As you keep on traversing, the application announces a series of suggestions thereby helping you reach a shopping mall, airport, or other transit stations. Apart from this, you can prepare a list of places you will visit, set routing options as per your preferences, explore the Street View option in Live mode, and so on. 

What is edge computing and why does it matter?

There are as many different edge use cases as there are users – everyone’s arrangement will be different – but several industries have been particularly at the forefront of edge computing. Manufacturers and heavy industry use edge hardware as an enabler for delay-intolerant applications, keeping the processing power for things like automated coordination of heavy machinery on a factory floor close to where it’s needed. The edge also provides a way for those companies to integrate IoT applications like predictive maintenance close to the machines. Similarly, agricultural users can use edge computing as a collection layer for data from a wide range of connected devices, including soil and temperature sensors, combines and tractors, and more. The hardware required for different types of deployment will differ substantially. ... Connected agriculture users, by contrast, will still require a rugged edge device to cope with outdoor deployment, but the connectivity piece could look quite different – low-latency might still be a requirement for coordinating the movement of heavy equipment, but environmental sensors are likely to have both higher range and lower data requirements – an LP-WAN connection, Sigfox or the like could be the best choice there.

Artificial Intelligence (AI): 4 novel ways to build talent in-house

To discover the gems hidden across your organization, you must start maintaining a self-identified list of skills for every employee. The list must be updated every six months and be openly searchable by associates to make it useful and usable. Palmer recommends self-classifying each individual’s skills into four categories: expert, functioning, novice, and desired stretch assignment. This allows teams with hiring needs to scout for individuals with ready skills and those with growth aspirations in the five competencies needed for AI. Finding the right content to upskill your in-house teams is a challenge. Despite the rapid mushrooming of training portals and MOOCs (massive open online courses), the curriculums may not meet your organization’s specific needs. However, with access to such great content online, often for free, it may not make sense to recreate your content. “You must design your own curriculum by curating content from multiple online sources,” says Wendy Zhang, director of data governance and data strategy at Sallie Mae. Base the training plan on your team’s background, roles, and what they need to succeed. 

Solving Mysteries Faster With Observability

Let's start by looking at the sources that we turn to when we look for clues. We often begin with observability tooling. Logs, metrics, and traces are the three pillars of observability. Logs give a richly detailed view of an individual service and provide the service a chance to speak its own piece about what went right or what went wrong as it tried to execute its given task. Next, we have metrics. Metrics indicate how the system or subsets of the system, like services, are performing at a macro scale. Do you see a high error rate somewhere, perhaps in a particular service or region? Metrics give you a bird's eye view. Then we have traces, which follow individual requests through a system, illustrating the holistic ecosystem that our request passes through. In addition to observability tooling, we also turn to metadata. By metadata, I mean supplemental data that helps us build context. For us at Netflix, this might be, what movie or what show was a user trying to watch? What type of device were they using? Or details about the build number, their account preferences, or even what country they're watching from. Metadata helps add more color to the picture that we're trying to draw.

Quote for the day:

"A sense of humor is part of the art of leadership, of getting along with people, of getting things done." -- Dwight D. Eisenhower

No comments:

Post a Comment