By Edoardo Russo, Lead Cloud Architect – Sense Reply
By Simone Casali and Francesco La Gatta, Business Developer Manager – Sense Reply
By Stefano Caliano, Business Unit Sr. Manager – Sense Reply
By Matteo Restelli, Sr. Partner Management Solutions Architect – AWS

Sense Reply

Regulatory reforms, renewable energy transition, e-mobility expansion, grid modernization, and the need to improve efficiency while reducing operational costs will push energy and utilities (E&U) companies to find new solutions and business models in the coming years.

Furthermore, many among oil and gas, as well as automotive, companies are shifting their business towards renewable energy production and electric vehicle charging. This expansion is accelerated through acquisition of plants, infrastructures, solutions and subsidiaries.

In these scenarios, large E&U organizations with centralized IT structures and enterprise architecture functions are called to serve a growing number of entities operating in several lines of business, facing the need to find a solution to reduce time to production and increase reliability.

In this post, we will discuss the main challenges that expanding companies face and how Sense Reply implements an Internet of Things (IoT) platform accelerator CI/CD suite on Amazon Web Services (AWS) to deliver an integration suite allowing project teams to develop in multiple accounts.

The Reply IoT Platform Accelerator (RIPA) is a ready-to-deploy, enterprise-level, and serverless IoT platform accelerator leveraging AWS services. It provides an automated, scalable, modular, expandable, and engineered portfolio of functionalities that supports heterogeneous business cases like energy and demand management, renewable management, smart grid and metering, e-mobility, and asset monitoring.

Reply is an AWS Premier Tier Services Partner and Managed Service Provider (MSP) with the many AWS Competencies, including IoT Consulting and Energy Consulting. An AWS Marketplace Seller as well, Sense Reply is part of Reply Group, which specializes on AWS as a whole process, with a focus on IoT for energy and utilities companies.


Releasing an IoT platform on multiple environments, and even on multiple accounts, may take a lot of time because a single IT department typically can’t hold all the requests coming from new teams onboarded in growing organizations.

The undifferentiated work of integrating new components and functionalities can subtract time and attention from serving business needs, and in some cases compromises on well-architected principles.

Sense Reply’s understanding of large E&U organizations highlights the following problems:

  • Slow time to market (weeks or months) is determined by centralized IT operations that don’t scale.
  • Focus on business functionalities gets lost in favor of rewriting functionalities in heterogeneous platforms.
  • Cost control, including licensing of third-party software.

Success can be achieved using an accelerator like RIPA, as partners help all entities in a complex organization incorporate the following principles in their projects:

  • Data standardization and centralization
  • Standardized device lifecycle management
  • Field multi-protocol integration
  • Security-by-design
  • Design for scalability
  • Repeatability
  • Ease of adoption for different subsidiaries
  • Reuse of business case agnostic components
  • Real time data propagation
  • Infrastructure as code (IaC)
  • Modularity
  • Adoption of pay-per-use services with upfront costs
  • Specialized for the typical use cases of E&U

Solution Overview

Sense Reply designed a solution to enable the integration of different vertical solutions for E&U businesses with a centralized set of functionalities made available by RIPA.

The goals of this suite are:

  • Give E&U organizations a tool that each team can use to manage only the infrastructure they need to run applications, developing the infrastructure definition as code, and let them focus on developing business logic to integrate with the core functionalities.
  • Grant resource segregation to facilitate cost sharing and accessibility between companies, teams, projects, and environments.
  • Keep track of changes for either single microservice and/or multi-microservices applications, enabling a detailed version control that allows easy management of the whole lifecycle of applications.
  • Provide a centralized account for the IT infrastructure team to store artifacts, source code, pipelines, and other fundamental resources as a shared service through the whole organization.
  • Allow agnostic deployment of applications either on the IoT edge or in the cloud.
  • Automate the creation of CI/CD resources, and preconfigure repositories with the help of templates that incorporate naming conventions.

Below is the high-level architecture of the Reply IoT Platform Accelerator CI/CD suite:


Figure 1 – CI/CD centralized architecture.

AWS CodePipeline for Microservices Architectures

When starting to organize the repositories with the above goals in mind, Sense Reply decided to divide all of the microservices applications in different repositories. As a result, the implemented microservices-oriented architecture has the following characteristics:

  • Supports small, independent, and decoupled microservices
  • Each microservice focuses on one feature
  • Faster and independent deployment and development
  • Technology independent
  • Version control on a single microservice
  • Easier coding of AWS CloudFormation templates
  • Standard naming convention

When a developer creates a new repository with the standard naming convention, AWS Lambda catches the event and prepopulates it. The developer will then find a predefined structure with some examples and a working pipeline ready to deploy the microservice in a development environment.

The pipeline, built on AWS CodePipeline, will manage the version control of the microservice and all of the interaction with the AWS CodeCommit repositories following the above guidelines.

In order to release a full application composed by many microservices in all of the other environments subsequent to Development (Test, QA, Production), Sense Reply developed a pipeline with the following characteristics:

  • Retrieve of all of the microservices infrastructure definition.
  • Application definitions (manifest files) are used to keep track of every change following some GitOps principles.
  • Aggregation of infrastructure definition.
  • Calculation of global application version based on a single microservices version.
  • Deploy only what has changed.
  • Promotion of artifacts and version control.

The following figure shows an example of an application lifecycle:

How code is propmoted through environments

Figure 2 – Environment pipelines flow.

Manage Lifecycle of a Serverless IoT Platform

To manage a serverless IoT platform in multi-environment and multi-account organizations, there are several principles to consider while developing the integration suite:

  • Segregation of resources: Implemented through naming convention and AWS tagging.
  • Rollback management: Possibility to decide which version of the core and the business line application to release.
  • Parametrization: Applied to every variable in order to adapt core resources to the destination accounts.
  • Account segregation: Segregation of NO-PROD and PRODUCTION accounts.
  • Environments isolation: Environments in the same account are completely isolated.
  • Security: Being aware of the best practice to not give full permissions to manage a project account from a shared account; each CloudFormation stack is run in the destination account.
  • Portability: All applications (core and no-core) can be treated as shared services and deployed on different accounts.
  • Version control: Every piece of code or resource must be versioned in order to track every change and enable the possibility to rollback.

In the end, every resource of the pipeline is written with CloudFormation templates, making it easily portable and easy to integrate with other AWS constructs to manage accounts, such as the account bending machine and the landing zone.

KPIs of the CI/CD Integration Suite

While using the integration suite in order to deliver the Reply IoT Platform Accelerator to clients, Sense Reply defined the following key performance indicators (KPIs):

  • Time to production for integration suite: 1 day
  • Time to production for RIPA on greenfield account: 2 days
  • Time to production for custom implementation integrated with RIPA: 4 hours
  • Creation of custom application pipelines: automated by the suite in 5 minutes
  • Creation of new pipeline builders: 4 hours

Instead of having to wait weeks or months for operations to deliver an IoT platform in a multi-account environment, it’s possible to have a fully functional RIPA environment in just a couple of days. It’s also possible to integrate new business solutions in just a few hours. It doesn’t matter if you have a “green” or “brown” field environment, the suite will adapt to the already present infrastructure, granting order in your IT teams and organizations.


Sense Reply, thanks to the CI/CD suite included in the IoT Platform Accelerator, addresses challenges faced by large energy and utilities companies, including slow time to market, loss of focus on business functionalities, and cost control.

The Reply IoT Platform Accelerator (RIPA) is a serverless IoT platform built on AWS, offering automated, scalable, and modular functionalities for diverse business cases in the energy sector.

Sense Reply advocates principles such as data standardization, centralized device lifecycle management, security-by-design, scalability, and infrastructure as code to achieve success in IoT platform integration.

The RIPA CI/CD suite adopts a microservices-oriented architecture using AWS CodePipeline, enabling independent and faster deployment of small, decoupled microservices with standardized naming conventions. The solution manages the lifecycle of a serverless IoT platform in multi-environment and multi-account settings, focusing on resource segregation, rollback management, parametrization, and version control.

Sense Reply defines KPIs for the integration suite, including rapid time to production for the suite and RIPA on greenfield and brownfield accounts, automated creation of custom application pipelines, and efficient pipeline builder development.

If you need support adopting methodology to manage large IoT solutions in complex E&U organizations, contact Sense Reply. You can also learn more about Sense Reply in AWS Marketplace.


Sense Reply – AWS Partner Spotlight

Reply is an AWS Premier Tier Services Partner and MSP with the many AWS Competencies, including IoT Consulting. Sense Reply is part of Reply Group, which specializes on AWS as a whole process, with a focus on IoT for energy and utilities companies.

Contact Sense Reply | Partner Overview | AWS Marketplace