Mauro Leonardo Sabatelli de Moura

I was the back-end developer of a web application that can do all the calculations needed to manage an investment fund.

It calculates portfolio holdings daily, based on previous and daily operations, performing asset and liability calculations. It automatically integrates with brokerage and custody accounts to retrieve all necessary information on day-to-day operations. It reconciles all data with fund administrators and other third parties involved in managing mutual funds, ETFs, and hedge funds.

The market data comes entirely from our internal data repository, updated daily from myriad data sources such as the Brazilian Stock Exchange (B3), Brazilian regulators (CVM, Anbima), and many other financial institutions.
The application also creates customized reports for back-office staff, fund managers, and fund clients. It also enabled further Power BI and Excel integrations by mapping its no-SQL database to an SQL database. It also communicates and alerts users using Discord API

Aside from the technical benefits, the company could eliminate $5+ thousand per month in external software licensing expenses by implementing a proprietary system.

A data repository of curated market information that can be used by financial portfolio managers and portfolio management systems. It extracts and curates asset information from several official data feeders and performs all necessary calculations, such as asset returns, volatility, and volumes, based on end-of-day prices, payouts, corporate actions, and all necessary information.

The application also analyses and loads macroeconomic data, company fundamentals from quarterly reports, and other specific information to allow users to extract data for their business intelligence reports.

The application has an automatic checking routine to detect missing or doubtful information based on many predefined algorithms to ensure the completeness and meaningfulness of the database.

I developed the back end of a call center app that kept a SQL database storing up-to-date pricing quotations for thousands of health insurance plans available in Brazil for small and medium businesses.

Call center agents could quickly determine the best price quotes for each customer by submitting the information required by the system through a REST API and then simulating the quotes according to the customer's needs.

The application stored the information about the simulations and the agents, in compliance with privacy regulations, LGPD and GDPR, for further analysis by the managers to increase the effectiveness of the sales team.

An E-Commerce platform based on dropshipping connects AliExpress sellers to Brazilian customers using the Shopify Platform. Facebook Ads were integrated into our website through Facebook Pixel. I have developed all the automations to properly create purchase orders to Ali Express and deliver the product to the customer's address using DSERS + Oberlo tools. The checkout+payment methods were performed with Yampi+MercadoLibre tools. Furthermore, I monitored the shipping status and integrated it into our website using 17Track API.

Developed an ML-driven application to read and parse structured information from a selection of approximately 50 email newsletters and RSS feeds using IBM Watson NLU. Keywords and sentiments were extracted from each message and a post-processing algorithm was able to obtain the most relevant topics and their sentiments from a given time period. The structured results were presented in a MySQL database that provided information for asset fund managers as well as for the company's weekly newsletter.

An automation app to automate and unify a workflow between the Pipefy and Monday apps using their REST/GraphQL APIs to integrate both systems into a central MySQL database using Node.js. Routines were run to synchronize data, and several automated scripts were triggered to create, upload, and download files across both systems and a data server. A dedicated relational database was developed to store and compute data, control and monitor script execution, and allow company managers to create custom reports to monitor KPIs and other custom indicators.

I've developed a complete Google Sheet automation using Google Apps Script to help users fetch data from external sources, validate input information, automate tasks and generate reports to further integrate with third-party apps.

It was created in a manner that also could be run in pre-scheduled routines and have access conditions to authorize users to perform controlled actions properly.

The AI Analyst Chatbot is an innovative, intelligent solution that provides comprehensive analytical support for businesses across various industries. Utilizing OpenAI API and the LangChain toolkit, this chatbot is adept at extracting specific data, answering complex questions, interacting with users, and generating in-depth analysis on relevant industry issues.

Developed in TypeScript with the NestJS framework, the chatbot is integrated with Discord to enable seamless user interaction, facilitating almost real-time communication and collaboration.
To manage and retrieve documents, the chatbot employs Pinecone, a vector database, allowing for efficient storage and retrieval of relevant information. Leveraging its intelligent document selection capabilities, the chatbot can identify pertinent documents and craft informed responses based on the user's request or query context.

This project allows companies to have improved decision-making through data-driven insights, reduced response time to critical inquiries, and enhanced overall operational efficiency.