- Delivered a 2-day training session on Prompt Engineering to a total of 24 employees. Each 8-hour session was divided between theoretical instruction and hands-on practice, allowing participants to develop use cases using the content learned.- Developed an automation that reduced 10 hours of manual work and generated $400 in monthly savings by retrieving admin cloud users from 4 different cloud providers (Azure, OCI, AWS, and GCP).- Developed an automation for user creation in Oracle databases, integrated with Jira, streamlining processes by validating database inputs, creating user accounts, sending credentials via email, attaching documentation, and automatically closing tickets. Achieved a cost reduction of $55 per month, impacting 19 tickets monthly.- Prompt engineering for LLMs

- Collaborated with clients in extensive meetings to gain a deep understanding of their business challenges and effectively translate their needs into tailored AI solutions.- Developed a tool using Embeddings (text-embedding-ada-002 model), GPT-3 API, Python and Streamlit to optimize the search for information in machine manuals. The main objective of this tool was to answer questions from machine maintenance workers accurately and efficiently, both in Portuguese and English. This could help them get information quickly without having to read the entire manual.- Developed a tool using GPT-3.5 API and Python to extract important topics from medical reports. The main objective of this tool was to optimize the work, since all topic extraction work was done manually by the company's employees.- Developed a tool using the Azure Speech to Text service, GPT-3.5 API, Python and Streamlit to extract relevant information about calls from a contact center, such as: reason for the call, customer feelings, whether the problem was resolved or not and the summary of the call. The main objective of this application was to allow managers to better evaluate the service and the main problems faced by their customers.- Developed a tool using LangChain's CSV Agent that helps users quickly extract insights from .csv files, without the need for programming knowledge.- Developed a tool that, using LangChain's Bing Search integration, makes it possible to use the entire content of a website as a knowledge base for a chatbot. In the project I developed, I directed the search exclusively to the website of the company I currently work for, but you can direct it to any website you want.- Developed a tool using the Gmail Toolkit, also offered by LangChain, to create an agent that aims to optimize the writing of emails. In this application, I implemented both a voice and text feature.

- Developed an automated minutes tool using Python, OpenAI (GPT-3 and Whisper), Reportlab and Streamlit, which guarantees the full transcription of the contents addressed, optimizes the workforce by reducing operational work and promotes greater focus on meetings and on the content being discussed.- Recorded a video for the company's YouTube channel on basic machine learning concepts- Wrote an article for the company's Blog about Voice Biometrics and its applicability - Wrote an article for the company's Blog about Voice Biometrics, deepening concepts

- Participation in payment flow projects, fixed asset control, among others- Consolidation and analysis of administrative expenses- Participation in other activities in the area