Recently, 13 states across the US placed a ban on the use of facial recognition technology by the police. Interestingly, 12 of these 13 cities were democrat-elect, implying the cultural difference within a country itself. The European Union is the gold standard when we talk about data privacy and laws governing the various aspects of technology. To protect individuals’ rights and freedom the article 22 of the GDPR, “Automated individual decision making, including profiling,” has ensured the availability of manual intervention in automated decision making in cases where individual’s rights and freedoms are affected. The first paragraph, “The data subject shall have the right not to be subject to a decision based solely on automated processing, including profiling, which produces legal effects concerning him or her or similarly significantly affects him or her,” and the third paragraph, “the data controller shall implement suitable measures to safeguard the data subject’s rights and freedoms and legitimate interests, at least the right to obtain human intervention on the part of the controller, to express his or her point of view and to contest the decision,” provides the right for manual intervention to individuals.
TPUs are Google’s custom-developed application-specific integrated circuits (ASICs) to accelerate ML workloads. A big advantage for GCP is Google’s strong commitment to AI and ML. “The models that used to take weeks to train on GPU or any other hardware can put out in hours with TPU. AWS and Azure do have AI services, but to date, AWS and Azure have nothing to match the performance of the Google TPU,” said Jeevan Pandey, CTO, TelioLabs. ... Google cloud’s open-source contributions, especially in tools like Kubernetes –a portable, extensible, open-source platform for managing containerized workloads and services, facilitating declarative configuration and automation– have worked to their advantage. ... Google cloud’s speech and translate APIs are much more widely used than their counterparts. According to Gartner’s 2021 Magic Quadrant, Google cloud has been named the leader for Cloud AI services. Pre-trained ML models can be instantly used to classify objects in an image into millions of predefined categories. Additionally, one of the top ML services from Google cloud is Vision AI, powered by AutoML.
RPA isn’t just a boon for patient-facing organizations—healthcare vendors are getting in on the action, too. For example, the company I work for faced the daunting challenge of transferring over 1 million pieces of patient data from one EMR to another. As any medical professional can attest, switching EMRs is a notoriously time-consuming process. So, we invested in RPA to bring efficiency to an otherwise manual and laborious task. In the end, we saved valuable time—and a significant chunk of change. ... One of the biggest contributors to burnout is the ever-increasing administrative work stemming from non-clinical tasks like documentation, insurance authorizations, and scheduling—all things that can be done faster and more accurately with RPA. And when providers are freed from the monotony, they have more time to focus on the parts of the job that they really enjoy. This, in turn, boosts morale and productivity, thus enhancing care delivery and optimizing patient outcomes overall. For those working in health care, the demand for digital solutions like RPA feels like the dawning of the new era—albeit one that is met with mixed emotions.
Managers in IT are sensitive to it, as complexity generally is their biggest headache. Hence, in IT, people are in a perennial fight to make the complexity bearable. One method that has been popular for decades has been standardisation and rationalisation of the digital tools we use, a basic “let’s minimise the number of applications we use”. This was actually part 1 of this story: A tale of application rationalisation (not). That story from 2015 explains how many rationalisation efforts were partly lies. (And while we’re at it: enjoy this Dilbert cartoon that is referenced therein.) Most of the time multiple applications were replaced by a single platform (in short: a platform is software that can run other software) and the applications had to be ‘rewritten’ to work ‘inside’ that platform. So you ended up with one extra platform, the same number of applications and generally a few new extra ways of ‘programming’, specific for that platform. That doesn’t mean it is all lies. The new platform is generally dedicated to a certain type of application, which makes programming these applications simpler. But the situation is not as simple as the platform vendors argue.
There is no precise definition of how big or small a microservice should be. Although microservice architecture can address a monolith's shortcomings, each microservice might grow large over a while. Microservice architecture is not suitable for applications of all types. Without proper planning, microservices can grow as large and cumbersome as the monolith they are meant to replace. A nanoservice is a small, self-contained, deployable, testable, and reusable component that breaks down a microservice into smaller pieces. A nanoservice, on the other hand, does not necessarily reflect an entire business function. Since they are smaller than microservices, different teams can work on multiple services at a given point in time. A nanoservice should perform one task only and expose it through an API endpoint. If you need your nanoservices to do more work for you, link them with other nanoservices. Nanoservices are not a replacement for microservices - they complement the shortcomings of microservices.
Knowledge of data-in-context, data processes, best techniques to provision, as well as tools enabling these methods of self-service are crucial to democratize data. However, with technology advancements, including virtualization, self-service discovery catalogs, and data delivery mechanisms, the internal data consumers can shop and provision for data in shorter cycles. In 2020, it took organizations anywhere between a week to three weeks to provision complex data that includes integration from multiple sources. Also, an increase in data awareness will help data consumers explore further available dark data that can provide predictive insights to create new user-stories that can propel customer journeys. ... A lack of focus is common across organizations as they assume Data Governance as an extension of either compliance or a risk function. Data Literacy will, in fact, change the attitude of business owners towards having to actively manage and govern data. There are immediate and cumulative benefits from actively governing data either by defining data or fixing bad quality data. But there is a need for a value-realization framework to actively manage the benefits of Data Management services.
Certifications are certainly important to consider when evaluating options, but even so, certifications don’t mean security. It is a best practice to check on the maturity of these vendor-specific certifications, as some companies go through a process of self-certification that doesn’t necessarily ensure the level of security your organization needs. Sending a thoughtful questionnaire to multiple vendors can be helpful for scoring these vendor’s security, offering a holistic and specific viewpoint to be considered by an organization’s IT team. On the customer end, in-house security and engineering staff can prep for CPaaS implementation by becoming familiar with the use of APIs and the authentication methods, communications protocols and the data that flows to and from them. Hackers routinely perform reconnaissance to find unprotected APIs and exploit them. Once CPaaS is incorporated into the hybrid work model technology stack, it is a best practice for an organization to focus its sights on its endpoint management. The use of a centralized endpoint management system that pushes patches for BIOS, operating systems, and applications is necessary for protecting the cloud network and customer data once a laptop connects.
Solution architecture is important, and yes, you want to minimize the number of daisy chains to reduce complexity. However, it doesn't mean you cannot have any daisy chains in your solution. In fact, dictating zero daisy chains can have consequences — not for performance, but for security. SASE consolidates a wide array of security technologies into one service, yet each of those technologies is a standalone segment today — with its own industry leaders and laggards. Any buyer who dictates "no daisy chains" is trusting that one single SASE provider can (all by itself) build the best technologies across a constellation of capabilities that is only growing larger. Being beholden to one company is not pragmatic given that the occasional daisy chain greatly increases the ability to unite best-of-breed technologies under one service provider's umbrella. ... SASE revolves around the cloud and is undoubtedly about speed and agility achieved through cloud-deployed security. But SASE doesn't mean the cloud is the only way to go and you should ignore everything else. Instead, IT leaders must take a more practical position, using the best technology given the situation and problem.
The work you’re doing as an SRE will partly depend on your company culture. Without a doubt, some organizations will relegate their SREs to driving existing processes like watching the on-call make sure there are no tickets, running deployments, etc. This can make folks feel like they aren’t progressing. However, today there are a lot more things you can do as an SRE than you once could. You used to just have Bash. Now you have many automation opportunities that will hone your programming skills. You can configure Kubernetes and Terraform. There's a bunch of code-oriented tools that you can use. You can orchestrate your stuff in Python. You could also use something Shoreline if you want it, which is “programming for operations,” and allows you to think of the world in terms of control loops, and how you can automate there. DevOps has also increased the Venn diagram overlap between SRE and Backend engineering. Previously, it was engineers using version control and engineers using package managers, which was separate from SREs using deployment systems and SREs using Linux administration tools.
When we need to change we usually feel a resistance against it. Take the current pandemic for instance. The simple action of wearing a facemask in public has caused indisputable resistance in many of us. Cognitively we understand that there is a benefit to doing so, even if there were long discussions on exactly how beneficial it would be. But emotionally it did not come natural and easy to most. Do you remember how it felt the first time you wore a facemask when entering the supermarket? It was not very pleasant, was it? But even when we are the driver for change we might find resistance against it. New year’s resolutions come to mind again. The majority of new year's resolutions are abandoned come February, even though the desired results have not been achieved. In other words, the resistance to change might sometimes show up late to the party. What might be missing here is endurance and resilience to small throw backs. I believe that we need a thorough understanding in which situation we currently are. This sounds simple and easy. And on a mid-level it is. "We need to come out of the pandemic with a net positive", a director of a company might say.
Quote for the day:
"It's very important in a leadership role not to place your ego at the foreground and not to judge everything in relationship to how your ego is fed." -- Ruth J. Simmons