Vera Varela

Data Scientist

Siemens

Apr 2026 - Present • 6 mos

Projects: • Next Best Action project: Determine the single most relevant product, contact, or service action for an individual customer, using statistics, artificial intelligence, or business rules. In this project: i) fix issues related to incorrect or missing data and change the frequency at which the NBAs are run; ii) add new details to existing NBAs. • Conversational AI Project: Lead the development of a chat available in SharePoint and MS Teams, designed to be used by more than 15,000 users to ask questions about information contained in documents, SharePoint pages, and the SharePoint calendar. In this project: i) coordinated with all the teams involved (Helix team, Arize team, SharePoint team) to manage access and permissions; ii) implemented the Thumbs Up/Thumbs Down functionality, which consisted of implementing changes in two GitLab pipelines, creating API keys in Microsoft Azure, and changing some configurations in SharePoint; iii) tested and fixed issues related to the bot’s access to SharePoint pages and the content within them, as well as its access to and retrieval of calendar content, ensuring that information such as layout, events, and times was displayed correctly, which consisted of using Arize to monitor the bot’s responses and identify whether the required information was missing from the retrieved chunks or whether the bot was providing an incorrect answer despite the information being available, using Postman to analyse the content in the retrieved chunks, and adapting the prompts in Helix; iv) held meetings with the Head of the department to share the status of completed and pending fixes, and communicated with the client via email and meetings to inform them about the features and fixes available for testing.

Data Engineer

Devoteam

Aug 2023 - Present • 3 yrs 2 mos

During a few months, help the client solve some issues in the Business Intelligence area. Currently, I'm working on a project to migrate a data warehouse (DW) from an on-premise environment to the cloud (GCP). My responsibilities include: i) developing data transformation pipelines using dbt (in Visual Studio Code); ii) deploying the project to the DEV environment through a CI/CD pipeline, where I run and validate the results; iii) once validated, promoting the project to the PRD environment. The entire process is version-controlled using GitHub.

Data Analyst

novobanco

May 2020 - Jul 2023 • 3 yrs 2 mos

Data Science projects: • NLP project (project developed alone) – Analyze the claims in the complaint portal to know the main reasons why customers complain. In this project: (i) I cleaned and treated the data, (ii) used BERT model (Hugging Face), (iii) used the UMAP technique, to reduce the dimension, (iv) used HDBSCAN algorithm, to create the clusters, (v) created a graph to see the distribution of clusters, and (vi) analyzed the results and prepared a report to present them. • Segmentation and Profiling project (project developed alone) – Analyse the types of corporate customers we have and the types of products those corporate customers bought, to identify which products we should propose to other corporate customers with similar characteristics. In this project: (i) I cleaned, treated and standardized the data, (ii) used decision tree algorithm, and (iii) analyzed the results and prepared a report to present them. • Machine Learning project – Predict the three best products for each individual client. In this project: After another team run a model, that used decision trees, where they introduced tens of variables and run the model for each product, that returned the importance of each variable for the specific product, (i) I analyzed the SHAP plots, generated by the model, to see the most relevant variables, their impact and correlation; (ii) analyzed if the most important variables was make sense, in terms of business; (iii) discussed with the team to take off or add some variable, considering the business; (iv) applied the conditions (extracted from the variables), in the data base, to find the best products for each client. Main tasks (as Data Analyst): Analyze the characteristics of customers and the products they have, with the purpose to create processes that help us to meet better each client, this way we can present the best products for each one of them (method used: Decision Trees).