Alejandro Figueroa

Served as technical lead and developer, architecting the solution and building its components for a cloud-based infrastructure designed to push real-time notifications to connected desktop applications. The solution aimed to provide a reliable and scalable way to send push notifications to a desktop application. These notifications needed to include details about real-time events generated by back-end services, such as new data available or changes made by other users in separate instances of the application.

Multiple back-end services published events to an Azure Event Grid topic, to be picked up by an Azure function and pushed to clients using Azure SignalR infrastructure. The desktop application connected to the SignalR endpoint using another Azure function to authenticate the request. The back-end services and client applications used reusable libraries published as NuGet packages to encapsulate the event generation and handling.

Built the solution as the main back-end developer and handled software design, construction, and deployment using DevOps practices. The RESTful API project was designed to capture and manage client requests for multiple business processes. Requests were composed of non-structured data, as each process has its own data requirements.

The RESTful API was implemented as an AWS Lambda function responding to HTTP requests. It received request data as JSON, along with some metadata, and saved it to an Amazon DynamoDB table. When a request changed status, the API sent a message to the SQS queue so it could be picked up by a processor AWS Lambda function using an event-based architecture. The processor function executed specialized business logic based on the request type.

The solution also included an Elasticsearch index to facilitate queries and searches. Whenever a request was created or updated, an Amazon DynamoDB stream sent a message that was picked up by another AWS Lambda function to update the Elasticsearch index. Deployment was made using AWS CodeBuild pipelines and AWS CloudFormation templates.

Built the application as the sole developer, working on both the front and back end. The application was a Vue-based web application to serve as a virtual agenda for kindergarten students, allowing parents and teachers to send messages and communicate about the daily routine of the children.

The front end was built as a Vue 2 application, composed of multiple reusable components, using framework features such as state management with Vuex and routing using Vue-router. Authentication was handled by Auth0 with custom workflow actions to handle authorization.

The back end was built using Azure functions as the API, backed with Azure Table storage as the data source. The database design followed an eventual consistency data model for integrity, creating redundant copies of the data tailored for specific query patterns. Whenever a new record was created or updated, the system sent messages to a queue to be processed by a separate Azure Function that handled data consistency.

Built a brand-new CI/CD pipeline as the main developer and author of the infrastructure as code components. The pipeline aimed to deploy three microservices using Docker containers to development, staging, and production environments.

The CI/CD pipeline was run on the company's on-prem Jenkins server and was triggered by successfully merged pull requests. The Jenkins job compiled the project, ran unit tests, and built the Docker images.

After that, the images were pushed to an AWS-hosted repository, and CloudFormation scripts were used to create the services in AWS Elastic Container Service (ECS)

Headed the project as architect and tech lead for a team of three developers. The solution was a RESTful API following the back end for the front-end pattern to support a brand-new mobile application.

This project involved creating a new RESTful API to support a brand-new mobile app, which was used by clients of the company to access its services. The API had to aggregate data coming from multiple downstream WCF services and expose them in a JSON schema to be consumed by the mobile app.

Headed the project as the tech lead for a team of three developers and participated in architectural decisions. The solution based on the BizTalk Server was built to automate the collections process of a financial services company. The project used BizTalk Server's messaging capabilities to automatically pull transaction data from banks and used XML maps to convert them to a canonical structure. It also used orchestrations and .NET libraries to execute the business process and send the data to the company's core system.

The solution exposed a web application where users could configure relevant messaging parameters such as the file format, connection parameters, and pulling frequency. The web app communicated with SOAP web services to save data to a database and trigger BizTalk Server orchestrations.