https://www.wine.com.br/ e-commerce website

In addition to normal routine procedures in a Scrum workflow, here are some projects that I've made key contributions on:

- BlackFriday 2021 page. Using Strapi and React.js we were able to make a very reliable experience during this campaign by using a simple JSON payload and so putting minimum stress on the backend while allowing for the needed flexibility for the sales teams to showcase the promotions.
- Web performance: applied the lazy loading technique to various images on the site's pages in order to improve the site loading experience. Performance is constantly measured with lighthouse/google page speed insights reports.
- Customer communication preferences central: I've created a page where users can choose which channels to receive company notifications. Later, I created a banner on the website's receipt screen so that users can opt-in to all channels again.

Some technologies/products I have contact with:
- React / Vue.js / Nuxt.js
- JS ES6 / Typescript / HTML5 / CSS3 / Sass
- Jest / Vue Test Utils / Testing Library
- Vuex / Mediator Pattern
- Atomic design
- Webback / Rollup / Gulp
- Figma
- Jira / Trello

First-level service (helpdesk). Treatment, classification, and formalization of demands in ticket system. Telephone and face-to-face service for users of the company; online and telephone assistance to Citizens and Information Technology managers in State agencies. User management of Government systems. Writing of knowledge bases.

Developed a public-facing searching engine for news articles related to some themes in politics. The job also involved a web and brand logo design for the website (Inkscape / Figma).

I've used the microservices architecture with the following technologies: web admin and API developed with Django; front-end site used Vue.js (with the Nuxt framework) where I used mainly the Element (UI), D3 (visualization), and ReactiveSearch (search components) libraries; the search backend used ElasticSearch and the data was stored in a PostgreSQL database. Many of the tasks in the backend were distributed and executed asynchronously via Celery Workers using the message broker RabbitMQ. For the website's contact page I've created a service for relaying the messages via e-mail with server-side JSON Schema validation and SPAM prevention using Google's Recaptcha, this was encapsulated in a small Node.js application.

The entire source of the project was versioned in Git using the feature branches model. It was deployed via Docker containers orchestrated with Docker Compose into a Linux VPS. To serve the requests a combination of a reverse proxy (Traefik), Nginx (to serve static files), and the application containers alongside a Memcached instance was used.

After some time in the approval environment, I received a change requirement arising from the costs of hosting an ElasticSearch search service. After searching, I've chosen to move the entire search to the client-side using the ElasticLunr.js library. Reengineering was carried out so that the backend would request the generation of the search index file from another service made in Node.js (Vercel / Micro) and finally this generated file would be consumed by the Nuxt client. The search in the frontend was implemented with customized components in Vue and a web worker.