Lucas de Medeiros

• Migrated over 50,000 students from a third-party platform to our internal system, preserving course progress despite the absence of structured APIs. Used Selenium to extract data and FastAPI with AWS Lambda to import and update student progress via internal APIs. All infrastructure was provisioned using Terraform.
• Automated charge cancellations by integrating Salesforce with our payment gateway via AWS. Built an API Gateway route that sends events to SQS, consumed by a Python-based Lambda that triggers the external cancellation API. This eliminated the need for manual actions by the sales team.
• Implemented student access control for our internal LMS product by developing API routes using FastAPI and the corresponding database models. This enabled integration with the contracts team for automatic and secure content release at scale.

Tech stack: Python, FastAPI, AWS (Lambda, SQS, API Gateway), Terraform, PostgreSQL, Selenium

• Implemented user profile management with personalized UI for each role (admin, student, teacher), removing the need to manage email-based access rules and significantly reducing front-end complexity. Used Kotlin with Spring Boot and PostgreSQL.
• Introduced an asynchronous job structure to automate student transfers between classes at year-end, eliminating manual processing and simplifying large-scale operations. Built with Java using Spring Boot ScheduledJobs.
• Refactored the teacher scoring feature using the Abstract Factory pattern and Spring's dependency injection, significantly improving code maintainability and simplifying future updates to business rules. Built with Kotlin and Spring Boot.

Tech stack: Java, Spring Boot, PostgreSQL, JPA.

• Migrated critical features from a Ruby monolith to decoupled microservices using Spring Boot and Kotlin, following Clean Architecture and DDD. Eliminated nearly 100% of flaky production errors and increased scalability with independent deployments. Microservices were hosted on AWS EKS clusters, using AWS SQS for event dispatch and retry management.
• Automated lead management and segmentation with an intelligent lead portfolio, replacing 5 manual spreadsheets. This doubled the number of leads analyzed in the final funnel stage. Used Kotlin for the backend, Next.js for the analyst's portfolio frontend.
• Implemented automation for credit data queries via REST API. Reduced lead processing time by 90%, cutting approval workflows from 30 min to less than 3 min. Used Kotlin and PostgreSQL for data storage.
• Designed business metrics monitoring with real-time dashboards and alerts in Datadog, sending custom metrics via Kotlin code. This improved funnel visibility, anticipating bottlenecks and potential system issues.
• Developed an event aggregation solution, consolidating data from multiple sources and sending it via Kafka to the Machine Learning team. This enabled lead-scoring algorithms and improved sales pipeline performance. Used Kotlin, PostgreSQL for storage, and Kafka for event delivery.

Tech stack: Kotlin, Spring Boot, AWS (EKS, SQS), PostgreSQL, Kafka, Datadog, Next.js