Articles

I am an « automation fundamentalist ». What I mean by that, is that I will go through large amounts of pain to avoid having to manually type code if I am able to have my computer do it automatically for me. The reasons are fairly self evident, because all bugs originates from a human being having written code at some point in time. Hence, if I can completely avoid manually writing code, the argument is that I can create 100% perfect software systems, that won’t even in theory be allowed to contain bugs of any kind. This idea extends to writing Unit Tests too.

Therefor, I created the ability to automatically generate Unit Tests in Magic. To understand how, watch the following video, where I demonstrate how I invoke an HTTP REST endpoint, for then to persist the invocation, allowing me to « replay » it later. The idea of course, is that being able to replay an HTTP invocation, can help me sanity check my system further down the road, as I modify it for whatever reasons I have to modify it.

Source de l’article sur DZONE

Robotic Process Automation, better known as RPA, is an exciting issue among the C-suite and is rapidly making progress across many businesses. Innovation has filled up rapidly in the past few years, as have debates regarding automation and other related breakthrough advances, including Artificial Intelligence.

The concept is expectable, considering how automation tools and technologies are closely linked. RPA has always been destined to be the future of automation. If you are continually feeling stressed over your business operations’ efficiency, utilizing RPA could be the ideal decision.

Top Robotic Process Automation Trends for 2021

In this section, we will explain some of the top RPA trends businesses should look out for in 2021 and beyond.

Source de l’article sur DZONE

In this sample, we will implement the server-side of a client-server style messaging app. To do so, you could use your preferred programming language, but for speed, we’ll use Linx, a low-code developer tool for backend APIs, integrations, and automation.

For a quick review of Linx and how it works, see this video.

Scope

We will implement some web methods, which will be useful for creating a messaging client. However, the following items fall outside the scope of this sample:

Source de l’article sur DZONE

As a consistent user and developer on the OpenShift platform over the years, I’ve tried helping users by sharing my application development content as we’ve journeyed from cartridges all the way to container base development.

With container based development we’ve also transitioned from using templates to define how to deploy our tooling and applications, to operators. There are many examples of how to work with the templated versions of our applications around decision management and process automation found on Red Hat Demo Central and JBoss Demo Central.

Source de l’article sur DZONE

PDF documents are small-sized, highly secure files. Almost all businesses use PDFs for processing their files. The reason being a distinguishing feature of maintaining format regardless of the tool used to access PDF files. It’s no surprise that all our invoices, official documents, contractual documents, boarding pass, bank statements, etc. are usually in PDF format.

Even as developers, we come across scenarios when a PDF file needs to be verified or used to locate certain parts of data. You can either do this manually given that you have loads of time to spare or you opt for automation testing. When it comes to handling tricky components of such files using automation, it might seem a bit too tricky. But that’s not the case. Selenium test automation can make it really easy to test PDF file formats.

Source de l’article sur DZONE

L’éthique est au cœur de la démarche IT de la MAIF. Robotiser les processus, oui, à condition que cela soit bénéfique aux sociétaires, aux collaborateurs, ainsi qu’aux performances et aux engagements sociaux de l’entreprise. À ce jour, une quinzaine de robots ont été déployés avec succès.

La MAIF est un assureur à part. Capitalisant sur ses valeurs mutualistes et sociales, la MAIF est financièrement solide, a décroché 16 fois le Podium de la relation client pour le secteur de l’assurance, tout en restant le seul assureur à faire partie des 50 entreprises les plus admirées des Français (IFOP – JDD – Nov 2019). Un impressionnant palmarès, mais qui ne doit pas faire oublier les questions d’efficacité opérationnelle. Un élément clé de satisfaction des sociétaires.

C’est en 2017 que la MAIF s’est penchée sur les technologies de RPA (Robotic Process Automation). Après une première expérimentation, l’assureur a compris tout l’intérêt qu’il pouvait tirer de cette technologie. Suite à appel d’offre, la solution de SAP a été choisie et déployée en 2018.

« Nous répondons aux sollicitations des métiers, qui expriment leurs besoins et décrivent une solution d’automatisation. Après étude du projet, le RPA est souvent préconisé pour répondre à ces demandes, car il est souple, facile à mettre en œuvre, avec un budget restreint » explique Jocelyn Paris, Chargé d’étude des systèmes d’information à la MAIF.

Avec le RPA, il n’est pas nécessaire de développer de nouvelles applications dans l’environnement MAIF : le robot va tout simplement reproduire des actions utilisateur. Ce qui est à la fois rapide et peu coûteux. Une quinzaine de robots sont aujourd’hui en production.

Une aide précieuse pour le service contrats

Les assureurs ont une responsabilité d’évaluation du niveau de risque de chaque assuré. Plusieurs fois par an, la MAIF fait ainsi ressortir les situations en alerte. Toutefois, un tiers de ces alertes peuvent être écartées selon des règles spécifiques et sans qu’il soit nécessaire de se mettre en relation avec l’assuré. Ce travail fastidieux peut être confié à un robot, libérant ainsi du temps pour les collaborateurs. Mis en place en un peu plus d’un mois, ce robot est un succès et se charge de sortir les fiches pouvant l’être de façon automatique.

Lors du batch d’avril 2020, les collaborateurs n’ont pas pu avoir un accès direct aux fiches papier. Le robot a donc été modifié pour reporter l’ensemble des fiches au prochain batch de septembre. Les services en charge d’imprimer et de traiter ces fiches ont été soulagés d’une tâche difficile à mettre en œuvre en télétravail. Les processus de RPA ont ainsi montré leur capacité à s’adapter rapidement à une nouvelle situation.

« Nous avons démarré avec un premier robot chargé de faciliter la régularisation des situations comptables de certains de nos assurés. Sur 15.000 cas recensés par an, 8000 à 9000 sont maintenant traités par ce robot, soit 1500 heures gagnées, témoigne Magalie Vincent, Responsable d’équipe au service contrats. Un nouveau robot, déployé en janvier 2020, aide à traiter les fiches d’observation de la sinistralité. Sur 8000 fiches, le robot a pu en écarter 2600. Cela simplifie notre travail en nous soulageant d’une tâche chronophage, qui mobilisait jusqu’à 15 collaborateurs. Nous pouvons ainsi nous pencher sur les cas les plus intéressants et nous recentrer sur notre rôle d’analyse du risque. »

Notez qu’une phase d’ajustement des critères d’action de ce nouveau robot devrait lui permettre de prendre en charge un plus grand nombre de dossiers.

Une accélération des rapprochements bancaires

Il est parfois difficile d’effectuer des rapprochements entre les paiements reçus et les contrats des assurés. Chaque jour, des centaines de règlements ne sont pas rattachés automatiquement au compte d’un sociétaire. La MAIF dispose d’un outil pour faire ressortir les virements non identifiés et d’un autre pour rechercher des coordonnées bancaires dans les bases de la société. L’affectation au bon sociétaire des fonds reçus sans référence reste toutefois un travail fastidieux.

Le rapprochement entre ces différentes bases de données est une tâche qui peut être en grande partie automatisée. Le robot se charge de prendre la référence d’un paiement dans le premier outil et de rechercher à qui il pourrait correspondre dans le second. Le service de la gestion économique dispose ainsi de plus de temps pour se pencher sur des dossiers complexes.

« Le robot nous aide dans ce travail fastidieux, long et très répétitif, confirme Sylvie Carasco, Chargée de gestion comptable, service comptabilité assurance et réassurance. Il est capable de rechercher simultanément dans plusieurs bases de données suivant des critères comme le nom ou l’IBAN, et de trouver un ou plusieurs sociétaires pouvant correspondre. Le travail de recherche est ainsi défriché, ce qui permet d’accélérer la clôture de chaque dossier. Ce robot est d’une aide précieuse, en particulier lors des périodes d’échéances annuelles, où 500 à 600 recherches doivent être effectuées chaque jour, occupant de 7 à 8 personnes à temps plein. »

Là encore, ce robot s’est montré particulièrement utile pendant la crise du Covid-19. Avec un taux de réussite de 80 %, il remplit ses objectifs. L’amélioration continue des critères d’action du robot devrait encore augmenter son efficacité.

Et des dizaines d’autres projets en préparation…

Le service qui gère les conventions automobiles en est déjà à trois robots déployés et un quatrième en cours de développement. Très intéressé par le RPA, le service conventions réfléchit à d’autres processus que cette technologie pourrait contribuer à automatiser.

« C’est un travail que nous avons mené sur plusieurs années, explique Céline Beunet, Responsable du service conventions. Les robots du service convention adressent des tâches précises : prise en charge d’un évènement non déclaré via l’ouverture d’un dossier et l’envoi d’un courrier au sociétaire ; automatisation du changement d’assureur lorsque le recours a été initialement présenté à la mauvaise partie ; suivi du remboursement du forfait, lorsqu’un recours présenté par une compagnie adverse a été contesté et refusé. Dans ce dernier cas de figure, le robot vérifie dans les échanges informatisés que nous avons bien été remboursés, puis classe le dossier. Ce que j’aime, c’est la possibilité de fixer les conditions d’action du robot et non de lui confier 100 % des processus. Ainsi, dès qu’un cas est litigieux, le gestionnaire reprend automatiquement la main. »

Le premier ROI du RPA est le temps qu’il fait gagner aux équipes. Économique et facile à mettre en œuvre, c’est un outil sans concurrence pour automatiser les processus. « Par rapport au coût d’automatisation d’une chaine logicielle, le RPA revient quasiment 10 fois moins cher, confirme Jocelyn Paris. Mais attention, le ROI est évalué selon quatre axes : la satisfaction des sociétaires ; la satisfaction des collaborateurs ; la performance de l’entreprise ; et l’aspect RSE. On ne veut surtout pas d’un robot qui remplace l’humain, mais d’une technologie qui aide l’humain. »

Maintenant que la plate-forme est en place et accessible largement, les demandes affluent et devraient sans nul doute s’intensifier. « La société française va vers plus de réglementation et d’administratif. Le RPA peut contribuer à atténuer ce phénomène. Si un robot peut s’occuper d’un traitement administratif, c’est aussi bien, car nous préférons que nos collaborateurs se focalisent sur leur cœur de métier : la relation humaine. » L’assureur militant s’engage ainsi au service de la performance comme du bien-être de ses collaborateurs.

The post RPA : la MAIF automatise ses processus, pour se concentrer sur l’humain appeared first on SAP France News.

Source de l’article sur sap.com


Automation

As engineers, we are always mindful of the need to minimize cost and time.  Experience has taught us that while automation is valuable in reducing code, we must also preserve design flexibility and enable agile iterations.  Let’s break that down a bit, and then illustrate with an example.

Automation (Reduce Routine Code — Executable Specifications)

To warrant the learning curve, automation must significantly reduce routine coding.  25%  reduction in code is not compelling enough.

Source de l’article sur DZONE

As a cloud-native developer you’ve installed an OpenShift Container Platform development environment on your local machine, but what’s next?

What can you do with the fully stocked container registry provided to you?

Source de l’article sur DZONE

Artificial intelligence. Just hearing the phrase has been a trigger for many in the technology world since that creepy Haley Joel Osment film circa 2001. But more recently, artificial intelligence and machine learning strike fear into the hearts of skilled workers for an entirely different reason: job security, or lack thereof.

Smart-home devices, streaming services, self-checkouts, even Google searches are ways that artificial intelligence has seeped into everyday life, exemplifying the abilities of computers and machines to master both simple and complex tasks. In some instances, these technological advancements make our lives easier, but for some people, their proliferation has meant job loss and skill replacement. There’s no wonder that when artificial intelligence starts being mentioned along with web design and site creation, the spidey senses of designers all over the world start tingling.

designers think outside the box, something that AI just can’t do

But let’s get real about what AI and automation really mean for designers for a second. Talented designers with busy schedules should view these advancements as virtual assistants. For some small businesses on a limited budget, the websites that artificial intelligence can pump out might be fine…for a while. However, as businesses grow, change, require updating and customization to adapt to their customer base, the expertise of creative and talented designers will always be needed. Even the best AI that we see today is limited by evaluating, replicating, and revising what already exists. It may be able to mix 1,000 different color schemes into 10 million potential combinations, but great designers think outside the box, something that AI just can’t do.

In fact, rather than being scared of automation, designers ought to embrace automation and artificial intelligence as a way to unleash their creative thinking. Delegate repetitive, straightforward tasks to the right software, and suddenly you have time to bring your best ideas to the table and push the boundaries of your own innovation. 

Where AI has Failed in Design

The ultimate goal of artificial intelligence and automation in design work is a grand vision that has yet to be realised.

Consider the case of The Grid, which began as a crowdfunding campaign in 2014. The “revolutionary” product posed itself as an artificial intelligence solution for building thoughtfully, yet automatically, designed websites in five minutes. Research “Reviews of the Grid” in any search engine and you’ll be met with scathing criticism with only some small praise sprinkled in. Most of the initial users cite underwhelming results, the feeling of being duped by the Grid’s marketing tactics, nonsensical placement of text, and ultimately, the Grid being a complete waste of money for the resulting product. Even at the low cost of $100, compared to hiring a talented designer, most users felt their investment was wasted.

For the AI capabilities that exist now, most small business owners, or those looking to put together a simple website, are better off using drag and drop site builders (Wix, Squarespace, Weebly, etc) that have been around for ages. Even so, there are plenty of businesses still willing to hire designers to take this simple task off their plate due to a lack of technical expertise or lack of time. And let’s be honest, are there even enough talented (keyword here!) designers out there to keep up with the millions of websites created every year, without each one working themselves to death? 

Where Automation Shines for Designers

Fortunately for good designers, it appears for now that the days of artificial intelligence completely taking over their jobs is a fantasy. However, what AI and automation do offer designers is a solid starting point for success, eliminating much of the lower-level grunt work that most designers would rather skip anyway.

Even well-received AI website builders like Firedrop still require a basic eye for design and specialised knowledge to produce truly unique, high-converting, and user-friendly websites. Tools and practices that designers should adopt are the artificial intelligence and automation resources that will help them do their jobs better, faster, and leave them with more time to focus on project elements that AI cannot accomplish on its own.

Bridging the Gap Between Designers and Developers

Well-established brands are likely to already have design systems in place that guide the creation of new elements across their digital profiles whether on social media, various mobile apps, or different sections of a website. But even in large corporations — excepting those who have perfected the process — there’s often a breakdown between a designer’s vision and resulting product from the developers. It stems from the basic difference in how they each approach their work and the limitations of the systems they use.

While component libraries — or even full design systems for that matter — won’t reconcile every question, they provide both developers and designers a source of truth to work from that both parties can understand. Design collaboration tools like Invision and Visme, specifically, keep designers and developers on the same page with automated version saving and code-friendly workflows.  

Understanding the Consumer

I don’t suggest using artificial intelligence to produce content for your site

Digging into and understanding the behaviours and habits of site users is a relatively new component of site design, but offers invaluable insights. Tools like HotJar, Mouseflow, or Smartlook make it simple to see holes or leaks in your conversion funnels, detect which page elements users are interacting with, and which they’re not interested in to refine the look and feel of a page for maximum conversions. Even though these tools provide the data, it still takes a keen eye and understanding of design to implement the right changes to improve site performance.

Site content is another way that artificial intelligence has the potential to improve our understanding of customer behaviour and improve site performance for individual users. I don’t suggest using artificial intelligence to produce content for your site, no matter how much the results have improved. However, static landing pages or a single set of further reading recommendations are unlikely to appeal to the majority of site visitors. Artificial intelligence tools like CliClap and Personyze instantly collect and analyse consumer data to provide dynamic, personalised experiences that drive more leads and encourage conversions. Creative designers will also learn from this data to improve customer experience with other pages or elements throughout the site.

Removing Distracting, Time-Sucking Administrative Tasks

Because “artificial intelligence” has become a term with such negative connotations, we often overlook the simple way that AI actually makes our work lives better and easier. Machine learning in email filtering is a great example of this. Consider a simple interface like a Gmail inbox. We have the option to mark certain senders as spam or as important, and our inbox learns that type of communication is and isn’t useful to the user. Pandora, Spotify, Apple Music, and more all take cues from the user behaviour of liking a certain song, artists, or genre of music to build customised playlists. There are a myriad of ways that artificial intelligence and its branches of disciplines merge with our everyday lives. 

Some of the most useful automations for business, and especially for designers, are related to the administrative tasks that frequently take time away or distract from more pressing projects. A perfect example of automation that can relieve stress and cut down on mindless work is an email autoresponder. I’ve always found that having time blocked off in my calendar to tackle complex or important projects helps me to focus on the task at hand and be more efficient. In order to more effectively block out my time, closing my email and setting an autoresponder to reply to all incoming emails serves two purposes: 

  1. Lets those trying to get in touch with me know that I only check my email at certain times of the day and that my response may not be immediate — tempering their expectations of when they might hear from me.
  2. Relieves my personal stress of being tethered to my inbox, splitting my focus, and also saves the time of having to initially respond to each email individually. 

This is just one simple way to use automation in your email, although there are many others to explore.

While Zapier isn’t the only workflow automation service on the market, it’s probably the most well known. Workflow automation reduces time spent on mind-numbing, repetitive tasks and helps designers connect apps that might not natively work together. Do you keep a task list in Todoist? Set up a Zap, then create a task in Todoist anytime someone mentions you on Asana or assigns you a task in Trello.

This is especially helpful for freelance designers who work with multiple clients across various project management platforms. The potential for automation to relieve unnecessary mental overhead for designers is nearly limitless.

Don’t be Afraid of AI, Embrace It

The bottom line of this brief overview of artificial intelligence and automation in design is that this emerging technology isn’t something designers should be scared of. In fact, it’s something to welcome with open arms because ultimately it can make our jobs, and our lives, better. Leave the monotonous tasks of collecting and analysing huge amounts of data or administrative minutiae to the machines; they can handle it.

Save the interesting, creative, abstract work for the talented designers who can turn AI recommendations into unique and intuitive digital experiences. Making the relationship between artificial intelligence and design symbiotic will yield the best results for every entity involved: the business, the AI, and yes, even the designer.

 

Featured image via Unsplash.

Source


Source de l’article sur Webdesignerdepot


Agile 

AI

Big Data

Cloud

Database

DevOps

Integration

  • Mulesoft 4: Continuous Delivery/Deployment With Maven by Ashok S — This article is a great example of what we want every tutorial to look like on DZone. The main aim of this article is to provide a standard mechanism to release project artifacts and deploy to Anypoint Platform, from the local machine or configure in continuous delivery pipelines.
  • Integration With Social Media Platforms Series (Part 1) by Sravan Lingam — This article helps you to build a RESTful API through MuleSoft that integrates with LinkedIn and shares a post on behalf of one’s personal account. I like this article because, in the age of social media, it’s so important for businesses to be connected and integrated!

IoT

Java

Microservices

Open Source

Performance

  • What Is Big O Notation? by Huyen Pham — Aside from a silly name, this article is an example of an in-depth analysis on a little-spoken-about concept. In this article, take a look at a short guide to get to know Big O Notation and its usages.
  • Is Python the Future of Programming? by Shormisthsa Chatterjee — Where is programming going? This article attempts to answer this question in a well-rounded way. The author writes, "Python will be the language of the future. Testers will have to upgrade their skills and learn these languages to tame the AI and ML tools".

Security

Web Dev

  • A Better Way to Learn Python by Manas Dash: There’s so many resources available for learning Python — so many that it’s difficult to find a good and flexible place to start. Check out Manas’ curated list of courses, articles, projects, etc. to get your Python journey started today. 
  • Discovering Rust by Joaquin Caro: I’m a sucker for good Rust content, as there’s still so many gaps in what’s available. Joaquin does a great job of giving readers his perspective of the language’s features in a way that traditional docs just 

Source de l’article sur DZONE