Renato – React, Typescript, AWS

An AI-powered recruiting platform that helps talent teams source, engage, and hire candidates at scale using a 200M+ candidate database, AI search agents, and automated outreach.

• Reduced manual screening time by up to 80% by building an AI “arbitrator” pipeline that unified sourcing, resume screening, and role-based ranking into a single automated workflow;
• Processed hundreds of thousands of LLM evaluations in under a minute by implementing parallel pipelines with prompt caching and extending Llama 3.1 via pre-training and fine-tuning for high concurrency;
• Enabled scalable automated talent search by developing a Python + FastAPI service orchestrating AI agents with LangChain, decoupling heavy AI workloads from the main app;
• Built and owned a full outreach automation system (templating engine, sequence editor, workflow manager, validation layer, and integrations), enabling personalized campaigns at scale;
• Drove credit-based revenue by implementing an AI prompt system generating personalized outreach using LLMs enriched with candidate and external data signals;
• Scaled infrastructure to support a 200M+ candidate dataset growing to billions of rows through database sharding and optimized data architecture;
• Improved performance by designing a dual-database setup using PostgreSQL and ClickHouse for high-throughput queries and RAG-based search;
• Accelerated recruiter workflows by building a dynamic content rendering system with variables, AI prompts, and reusable snippets;
• Delivered real-time updates during large-scale imports by building a processing pipeline integrating ATS, parsing resumes with Apache Tika and Tesseract;
• Improved system consistency by redesigning the GraphQL API for better reuse, deduplication, and domain composition;
• Reduced errors and saved time by implementing a reusable snippets system with fine-grained permissions;
• Ensured scalability and maintainability by architecting a modular Node.js + Bun monolith with clear domain boundaries and event-driven architecture.

An analytics product within a trade data network that provides real-time visibility into trading and clearing costs, enabling firms to reconcile exchange fees, clearing commissions, and brokerage in a single platform.

• Scaled to thousands of daily transactions from top firms like BlackRock by designing RESTful APIs to support a new line of business;
• Launched a brand-new product in under a month by working closely with non-technical stakeholders across departments to define and scope requirements;
• Enabled shared component reuse across packages by redesigning the frontend into a monorepo using Yarn Workspaces, extending an existing product and its design system;
• Cut the release cycle time by 10x by restructuring CI/CD pipelines from feature-branch to trunk-based development.

A healthcare technology company that combines care navigation with a transparent pharmacy benefit manager (PBM), pairing members with clinical guides through a mobile app to help them find high-quality care and maximize their benefits. It serves over 30 Fortune 500 companies and 4 million+ members.

• Made cross-team alignment significantly more efficient by introducing DDD with well-defined business domains and ubiquitous language to structure architecture discussions;
• Produced a leaner backlog and improved user experience outcomes by partnering with the product team to rethink feature scope and minimize overengineering;
• Reduced ambiguity and onboarding time for new engineers by laying the React architectural foundation with structured, layered definitions across the app;
• Improved reliability and consistency across external integrations by establishing a gRPC communication design for synchronous service operations in Node.js;
• Kept domain logic free from database-specific concerns by designing a thinner database layer using Knex and PostgreSQL, replacing repositories with direct application service queries;
• Reduced integration test time from over 5s to under 1s per test by leading performance improvements using React Testing Library and Vitest;
• Enabled horizontal scalability and infrastructure cost reduction by restructuring the backend architecture on top of NestJS microservices;
• Enabled the customer service team to process thousands of client files per day by introducing a CMS that reduced new client onboarding friction;
• Cut CSV processing time by roughly 10x by implementing parallel stream processing in Node.js and PostgreSQL for files with millions of rows;
• Improved scalability and reduced inter-service dependencies by establishing an event-driven architecture using CQRS and Kafka for asynchronous cross-system operations.

An insights management platform trusted by all the top 20 pharmaceutical companies that unifies field activity, HCP engagement, social sentiment, claims data, and congress monitoring into a single, decision-ready view of the market.

• Extended existing functionality without breaking data compatibility by redesigning PostgreSQL schemas with migration backtrack strategies and deprecation methods;
• Reduced UI coupling and improved frontend maintainability by refactoring the architecture to eliminate component nesting and prop-drilling through composition over inheritance and state collocation;
• Significantly improved test performance and code clarity by migrating from Enzyme to React Testing Library and introducing BDD with the Testing Trophy.

As part of a People Tech company, contributed to two projects: a leading Brazilian credit recovery and financial services provider with 25+ years in the market and 17,000+ employees, and an award-winning digital consultancy specializing in enterprise ecommerce, CPQ, and data management for global brands.

• Processed millions of financial transactions per day in real time by designing a billing CRM platform with an event-driven architecture;
• Aligned engineering and product around a shared domain vision by introducing Domain Storytelling and building Context Mapping with Bounded Contexts and strategic DDD patterns;
• Met the high customization demands of major banking clients such as Itau and Santander by designing a distributed multi-tenant system with physical tenant isolation;
• Achieved sustainable and resilient physical data isolation across tenants by designing distributed data systems using clustered PostgreSQL in a multi-tier architecture;
• Reduced over-fetching and improved UI performance by designing a BFF service as a Federation API Layer in Node.js using a DataLoader architecture with GraphQL and Nexus;
• Increased development speed and reduced integration delays by adopting Trunk-Based Development with Devspace.

A Brazilian SaaS legaltech company providing contract lifecycle management (CLM) software that streamlines contract creation, negotiation, and management through document automation and customizable workflows for medium and large enterprises.

• Guided the engineering team to replace a legacy system by building a new platform from scratch;
• Built a frontend architecture with GraphQL client integration for incremental data updates;
• Introduced React component standard definitions to establish consistent scaling patterns;
• Built a top-level form API using Formik and Yup;
• Configured Jest and React Testing Library processes to reduce test maintenance friction;
• Co-led efforts to expose and stabilize public-facing APIs.

A Brazilian fintech operating a digital lending platform focused on asset-backed consumer loans (homes, vehicles, payroll), offering lower interest rates and expanding into insurance, a car marketplace, e-commerce, and employee benefits.

• Established a mobile engineering capability from the ground up by building the team from scratch through external hiring and internal recruiting;
• Aligned engineering efforts with business priorities by partnering with business experts to define and set yearly OKRs across company products;
• Enabled consistent mobile development across all cross-functional squads by redesigning the mobile engineering project for scalability and shared maintainability;
• Ensured data model consistency across business domains by contributing to a platform restructuring with improved data-fetching strategies;
• Balanced refactoring with feature delivery without disruption by managing continuous-delivery releases using feature flags;
• Increased mobile app delivery to multiple releases per day by optimizing CI/CD pipelines;
• Reduced customer support response time by at least 10x by introducing impersonation mechanisms in development workflows using production data, enabling more precise engineering support.

A venture builder that develops custom apps and software for large enterprises, where I worked on projects for a global cosmetics multinational and a major airline operating extensive regional and international routes.

• Introduced practical TDD and provided technical assistance via pair-programming and mentorship;
• Established testing strategies for Relay.js to improve test maintainability;
• Isolated database integration mocks for Jest SQL query tests to reduce CI CPU costs;
• Reported engineering KPI outcomes to business stakeholders;
• Co-led technical autonomy initiatives to reduce dependency on senior engineers;
• Worked directly with clients to understand and solve problems;
• Built a data-driven frontend architecture with Relay and Redux integrated with a Node.js aggregation layer served by Koa over GraphQL;
• Managed multiple workers for parallel processing in Node.js;
• Shipped iOS/Android apps with Fastlane; shipped Node.js services with Docker and Kubernetes on AWS; shipped SPAs with Nginx on Amazon CloudFront.