Per prima cosa cominciamo a definire i requirements…

• Un numero telefonico dedicato a WhatsApp Business API (Può essere benissimo un numero telefonico Virtuale ma deve supportare integrazione con WhatsApp).
• Un Webhook al quale WhatsApp invia al backend dell’utente i messaggi inviati dall’interlocutore.
• Una logica di invio dei messaggi da parte dell’utente all’API di Meta.

L’implementazione delle WhatsApp API parte dal portale sviluppatori di meta: https://developers.facebook.com

Qui lo sviluppatore deve accedere con i suoi dati di accesso a facebook, o con i dati di accesso della sua impresa o alternativamente, se lavora per un’impresa, questa deve condividere le risorse di amministrazione del profilo WhatsApp API con il profilo facebook dello sviluppatore.

Una volta effettuato l’accesso bisogna creare una nuova App.

Aggiungiamo un nuovo numero di telefono e lo autentichiamo (riceveremo un messaggio contenente un codice da inserire nella sezione developers).

Io consiglio sempre di avere un file di enviroment variables (.env) protetto da scrittura e lettura pubblica incapsulato in modo che solamente il backend possa accedervi attraverso un metodo dedicato.

Sviluppando in PHP vi consiglio la libreria composer vlucas/phpdotenv

Si richiama all’interno della classe in questo modo:


use DotenvDotenv;
$dotenv = Dotenv::createImmutable(__DIR__);
$dotenv->load();


Una volta registrato il numero dobbiamo recuperare questi parametri:

WHATSAPP_VERIFY_TOKEN={un token che creiamo noi arbitrariamente ma che ci servirà per iscriverci al servizio}
WHATSAPP_ACCESS_TOKEN=(il token staccato dal developers portal di facebook)
WHATSAPP_PHONE_NUMBER_ID=(l’id che facebook associa al nostro numero di telefono precedentemente registrato)
WHATSAPP_NUMBER=(il numero di telefono che abbiamo registrato)

Una nota importante… Il numero di telefono registrato su WhatsApp Business API non può essere poi utilizzato anche per utilizzare WhatsApp su Cellulare.
L’utente a quel punto può scegliere se memorizzare i messaggi su database per rispondere attraverso un’interfaccia Web oppure se aggiungere un LLM addestrato che si occupi di rispondere autonomamente all’interlocutore.

In questo articolo vedremo proprio questa casualità.

Una volta ottenuti tutti questi dati, dobbiamo andare a inviare il nostro primo messaggio verso una numerazione telefonica (ad esempio il nostro numero personale).

Nella console di Facebook developers bisognerà recarsi nel menù WhatsApp -> Configurazione API (dove abbiamo creato il nostro token).
Apro qui una breve parentesi. Il token di accesso staccato da questa interfaccia è un token a scadenza della durata di 60 giorni. Se si vuole generare un Token dalla durata illimitata bisogna registrare l’utente come Utente di Sistema da business.facebook.com concedergli le risorse di amministrazione per l’app che abbiamo creato in developers e staccare un token con scadenza illimitata.

Da qui inseriamo il nostro prefisso telefonico e il numero del nostro cellulare personale dove vogliamo ricevere il primo messaggio e cliccando su Invia Messaggio se tutto è stato impostato correttamente dovremmo ricevere il messaggio WhatsApp.

In basso, al passaggio 2 facebook fornisce anche una istruzione cURL per l’invio dei messaggio che possiamo riutilizzare per creare un metodo di invio dei messaggi con cURL verso l’endpoint di WhatsApp API.

Abbiamo gestito la parte in cui siamo noi ad inviare un messaggio. Dobbiamo gestire la parte in cui è l’utente a mandare un messaggio a noi.

Ci spostiamo su Webhook e nella select selezioniamo come Product: Whatsapp Business Account

Qui dovremo inserire il nostro URL di Callback che corrisponde all’URI del nostro Webhook.

Come token di verifica inseriamo il token che abbiamo creato nella variabile d’ambiente WHATSAPP_VERIFY_TOKEN nel file .env.

Cliccando su verifica e salva WhatsApp chiamerà la nostra API per verificare che il Webhook sia attivo e funzionante ed iscriverlo correttamente al servizio.

Se vogliamo osservare cosa ci sta inviando WhatsApp consiglio di utilizzare tecniche di reverse-proxy, ad esempio utilizzando Ngrok e sostituendo l’indirizzo del Webhook con quello fornito da Ngrok.

Il Webhook è il punto in cui avviene la magia. Possiamo limitarci a scrivere i messaggi ricevuto in un database con un timestamp, oppure possiamo istruire una rete neurale in OpenAI per far rispondere un LLM alle domande dell’utente per le quali il modello custom trained è stato istruito. A seconda della natura del tema di cui dovrà occuparsi è possibile anche utilizzare i modelli nativi di OpenAI (per compiti non troppo impegnativi consiglio il modello 4o-mini che è il meno costoso. Il costo è di circa 0.15 euro per milione di Token di Input e di 0.60 euro per milione di Token di output alla data odierna. In una semplice chat si arriva a spendere un euro dopo migliaia di messaggi scambiati considerando una media di 500 token totali per ogni richiesta.

Non abbiamo ancora parlato dei costi di WhatsApp…

Per le chat aperte dai clienti, il servizio è completamente gratuito in una sessione di 24 ore. In WhatsApp infatti le sessione che vengono aperte hanno una durata di 24 ore prima di scadere. Se siamo noi ad aprire una chat verso un cliente il costo per sessione è di circa 5 centesimi di euro per sessione aperta verso un numero univoco. Una volta aperta la sessione non ha importanza il numero di messaggi inviati. WhatsApp conteggia unicamente la sessione.

Per fare un esempio pratico… Se un nostro cliente richiede informazioni al nostro bot alle 13:00 del 12 aprile 2025 è lui che sta aprendo una sessione che sarà gratuita e che scadrà alle 12:59:59 del 13 aprile 2025. Se risponderemo al messaggio dopo questo orario apriremo noi una nuova sessione che diventa a pagamento (5 centesimi di euro) per poter scambiare tutti i messaggi che vogliamo fino alle 24 ore successive. Se al termine delle 24 ore invieremo un ulteriore messaggio verrà aperta una nuova sessione al costo di 5 centesimi.

In buona sostanza il costo delle conversazioni su WhatsApp Business API dipende unicamente da chi sta aprendo la sessione. Un cliente? Costo zero. Noi che sfruttiamo le API? 5 centesimi a sessione di 24 ore.

Sulle API di WhatsApp è inoltre possibile avviare delle procedure guidate, ad esempio è possibile identificare l’utente che sta richiedendo informazioni attraverso un form di login. Questa procedura sfrutta i WhatsApp Flows che saranno argomento di un articolo dedicato.

[starbox] 

L'articolo WhatsApp Business API proviene da RENOR & Partners S.r.l..

Google Cloud Run and AWS Lambda are two of the most prominent serverless technologies gaining traction.

In this article, we’ll examine both platforms, comparing them across key categories such as pricing, performance, programming language support, scalability, and more.

Introduction to Serverless Technologies

Serverless technologies are cloud platforms that automatically scale without the need for dedicated server infrastructure. They also allow developers to rapidly build and deploy functionalities, saving time on server configuration and maintenance.

Serverless platforms support horizontal scaling, ensuring high performance during traffic spikes.

Google Cloud Run Description

Google Cloud Run is Google’s serverless container platform that allows developers to deploy an entire backend infrastructure without managing a server.

By leveraging Docker containers, Google Cloud Run enables developers to build, deploy, and integrate their serverless functions with ease.

In addition, Google Cloud Run supports integration with other Google Cloud Platform (GCP) services such as Cloud Storage for storage, OAuth 2.0 for authentication and authorization, Cloud Identity for user management, Cloud Load Balancing for networking, and more.

Description of AWS Lambda

AWS Lambda is a serverless platform from Amazon Web Services that allows developers to deploy code without configuring a server. It also enables code execution in various environments such as Node.js, Java, C#, Ruby, Python, Go, and more.

AWS Lambda is also a microservices platform that enables the creation of scalable applications. You can use AWS Lambda to run code in response to events such as HTTP requests, database triggers, or system notifications.

Price comparison

Google Cloud Run offers pay-as-you-go pricing based on the number of containers run and their execution time. The base price for running a container is $0.0025 per minute, with a maximum rate of $0.10 per minute.

AWS Lambda, on the other hand, offers pay-as-you-go pricing based on the number of requests and execution time. The price for requests is $0.20 per million requests, with a maximum rate of $0.25 per million requests.

Performance Analysis

Google Cloud Run offers high scalability and fast response times. With Cloud Run, containers run on a single instance of Google virtual machines, providing high processing speed.

AWS Lambda, on the other hand, can automatically scale to handle millions of requests at once, offering faster response times. Additionally, AWS Lambda supports up to 3.2 GB of memory per instance.

Programming Language Support

Google Cloud Run supports a wide range of programming languages, including Java, Node.js, Go, Python, Ruby, and PHP. It also supports additional languages such as C#, C++, Rust, and more.

AWS Lambda also supports a wide range of programming languages, including Java, Node.js, Go, Python, Ruby, PHP, and many more. It also supports environments such as Node.js, the Java Virtual Machine (JVM), and the Common Language Runtime (CLR).

Scalability and Availability

Google Cloud Run offers automatic horizontal scalability to handle real-time requests. It also supports high availability to ensure that applications remain consistently up and running.

AWS Lambda, on the other hand, is designed to automatically scale based on incoming requests and offers high availability with minimal downtime.

Serverless Function Development and Deployment

Google Cloud Run provides an end-to-end development solution for creating serverless functionalities. It offers features such as building, testing, editing, pushing, deploying, monitoring, and configuring containers.

AWS Lambda, on the other hand, provides developers with tools to easily create, test, and deploy serverless functions. You can use AWS Lambda to run code in response to events such as HTTP requests, database triggers, or system notifications.

Google Cloud Run and AWS Lambda are both related to serverless technologies and offer developers a variety of tools and features to build scalable applications. However, Google Cloud Run and AWS Lambda differ in terms of pricing, performance, programming language support, scalability, and more.

Before choosing one of these platforms, it’s important to carefully evaluate your own requirements and decide which one best fits your needs.

[starbox]

L'articolo Comparison: Google Cloud Run vs AWS Lambda proviene da RENOR & Partners S.r.l..

Google Cloud: the solution for business efficiency

Google Cloud offers a secure and reliable infrastructure for data storage and business service management. Thanks to its ability to scale flexibly, companies can reduce costs and increase productivity. In addition, Google Cloud provides advanced tools for monitoring and managing data, with the ability to access information from anywhere.

Improving collaboration between departments with Google Cloud

Thanks to Google Cloud, company departments can collaborate more efficiently and in a more coordinated way, improving communication and information sharing. Google Drive, for example, provides a collaborative environment that allows team members to share documents, files, and other data in real time. This reduces working time and increases productivity.

How to manage projects more efficiently with Google Cloud

Google Cloud offers a wide range of tools for project management, such as Google Calendar, Google Keep, and Google Tasks. These tools allow you to organize work more efficiently, track tasks and deadlines, and collaborate with other team members. In addition, Google Cloud enables access to data from anywhere, making project management even more flexible.

Optimizing business processes with Google Cloud: here’s how

Google Cloud offers a wide range of tools to optimize business processes, such as human resource management, accounting, and inventory management. Thanks to its ability to integrate with other applications and services, Google Cloud allows for the automation of processes and the reduction of working time. In this way, companies can become more efficient and productive.

The advantages of Google Cloud for data security and privacy

Google Cloud is designed to ensure the security and privacy of business data. Thanks to advanced security tools, such as data encryption and two-factor authentication, companies can protect their data from unauthorized access and loss. In addition, Google Cloud provides regular security updates and a team of experts dedicated to data protection.

Google Cloud: the ideal choice for business efficiency and productivity

In summary, Google Cloud is an ideal solution for companies looking to improve their efficiency and productivity. Thanks to its reliability, flexibility, and security, Google Cloud offers a wide range of tools for managing data and business processes. In addition, Google Cloud is constantly evolving, with new features and services being added regularly. If you want to make your company more efficient and productive, consider Google Cloud as your solution.

[starbox]

L'articolo How to use Google Cloud to improve business efficiency proviene da RENOR & Partners S.r.l..

GitHub Copilot: what it is and how it works

GitHub Copilot is an AI-powered programming assistant developed by GitHub and OpenAI. Using the OpenAI GPT-3 API, Copilot can automatically generate code, suggesting it to the user based on the context in which they are writing the program.

VS Code and GitHub Copilot: a winning combination

GitHub Copilot was developed to be integrated with Visual Studio Code, Microsoft’s popular open-source code editor. This integration creates a powerful development environment capable of suggesting, completing, and generating code dynamically, supporting the user throughout the programming process.

A virtual assistant for coding

GitHub Copilot works as a virtual assistant for coding, suggesting and completing code based on the user’s current context. Copilot can generate complex code, such as functions, classes, and interfaces, and can also complete parts of already existing code.

Machine learning at the service of developers.

GitHub Copilot is based on artificial intelligence and machine learning, which allow it to learn from existing code and generate new code with high precision. Copilot uses the OpenAI GPT-3 API, which enables it to process a vast amount of data and generate code with increasing accuracy and detail.

Practical examples of using Copilot

A practical example of using GitHub Copilot could be writing a function that converts a string into a date. Copilot would suggest creating the function with the correct format and appropriate arguments, completing the implementation with the necessary code. In this way, the user would save time and effort during the programming process.

Advantages and limitations of AI in code writing

The advantages of using AI in code writing are undeniable: a higher level of automation, greater accuracy, and faster development speed. However, like any technology, AI also has its limitations: the user must be able to understand the generated code and correct it if necessary.

Privacy and data security on Copilot

An important aspect to consider when using GitHub Copilot is data privacy and security. Copilot is powered by a large amount of data, including users’ code and personal information. However, GitHub has stated that Copilot respects user privacy and that the data is encrypted and stored securely.

The future of coding: AI as a support tool

Artificial intelligence is set to become increasingly important in the world of programming. GitHub Copilot is just one of the first examples of how AI can be used as a support tool for developers. In the future, we are likely to see more and more programmers using AI to automate parts of their work and improve their efficiency.

In conclusione, GitHub Copilot rappresenta un importante passo avanti nella programmazione assistita dall’AI. Sebbene ci siano ancora limiti da superare, l’utilizzo dell’intelligenza artificiale nella scrittura del codice potrebbe portare a un maggiore livello di automazione e di efficienza per i programmatori. Con l’integrazione in Visual Studio Code, l’utilizzo di Copilot potrebbe diventare ancora più semplice e immediato, permettendo a molti sviluppatori di migliorare la loro produttività.

[starbox]

L'articolo GitHub Copilot: the perfect assistant for VS Code proviene da RENOR & Partners S.r.l..

Serverless cloud: what it is and how it works

The concept of serverless cloud refers to the ability to build and manage applications without having to manage a physical server. In simple terms, serverless allows application code to run directly in the cloud, without the need for a dedicated server. This way, serverless helps reduce costs and simplifies the management of IT resources.

Google Cloud Run: Google’s solution

Google Cloud Run is a serverless computing solution that allows applications to run quickly and flexibly without having to manage server resources. Cloud Run enables Docker containers to run automatically, on demand, and with no configuration costs. In this way, Cloud Run allows applications to run without worrying about infrastructure management, simplifying developers’ lives.

Advantages of using Google Cloud Run

Using Google Cloud Run offers numerous advantages. First, it helps reduce IT resource management costs by eliminating the need for a dedicated server. Second, it allows applications to run quickly and flexibly, without the need to worry about server configuration. Additionally, Cloud Run enables the creation of scalable and secure applications, consistently ensuring top performance.

How to use Google Cloud Run

To use Google Cloud Run, you need to create a Docker image containing the application to be executed. Once the image is created, you can run the application on Cloud Run by selecting the desired options and configuring the service. Once configured, Cloud Run allows the application to run automatically and without any setup.

Create your own serverless application with Google Cloud Run

To create your own serverless application with Google Cloud Run, you need to create a Docker image containing the application to be executed. Once the image is created, you can run the application on Cloud Run by selecting the desired options and configuring the service. Once configured, Cloud Run allows the application to run automatically and without any additional setup.

Manage resources and services with Google Cloud Run

Managing resources and services with Google Cloud Run is simple and intuitive. Cloud Run allows you to configure desired options and monitor the execution of applications precisely and in detail. Additionally, Cloud Run enables scalable and secure management of resources and services, always ensuring optimal performance.

Examples of applications built with Google Cloud Run

Google Cloud Run can be used to develop a wide range of applications, including web applications, mobile applications, and IoT applications. Some examples of applications built with Cloud Run include data management applications, cybersecurity applications, and IT resource management applications.

Improving performance with Google Cloud Run

To improve application performance on Cloud Run, you can use advanced caching and data caching features. Additionally, CDN services can be used to enhance application speed and scalability.

Security and scalability with Google Cloud Run

Google Cloud Run ensures the security and scalability of applications through advanced security features and IT resource management. Additionally, Cloud Run allows applications to be managed in a scalable and secure way, ensuring high performance and sustainable costs.

Costs and pricing of Google Cloud Run

Google Cloud Run offers a flexible and cost-effective pricing model that allows you to pay only for the resources you use. In addition, Cloud Run provides a wide range of pricing options, including pay-as-you-go and package-based pricing models, enabling you to choose the solution that best fits your company’s needs.

In summary, Google Cloud Run is a reliable and flexible solution for building serverless applications. Thanks to its advanced features, Cloud Run allows applications to be developed quickly and flexibly, without the need to manage IT resources. Furthermore, Cloud Run ensures the security and scalability of applications, consistently delivering top performance and cost efficiency.

[starbox]

L'articolo Serverless cloud with Google Cloud Run proviene da RENOR & Partners S.r.l..