Alan Marcell de Oliveira

AI-Driven Engineer → Applied AI Engineer

alanmarcell.github.io	alanmarcell@live.com
github.com/alanmarcell	linkedin.com/in/alanmarcell

Professional Summary

Software Engineer with nearly 10 years of experience building high-performance backend systems, scalable APIs, and cloud solutions, now focused on Applied AI: designing systems where LLMs are a runtime component, not just a coding accelerant. I build and ship operational AI agents, tool-calling workflows, and agentic architectures in production, on top of solid engineering fundamentals in distributed systems, event-driven architecture, and cloud infrastructure. This combination turns AI from a productivity trick into a strategic capability that solves real business problems.

Technical Skills

Languages & Core: Node.js, TypeScript, JavaScript, GraphQL
Cloud & Infrastructure: AWS (Lambda, SQS, SNS, Step Functions), Serverless Framework, Azure, Terraform, Docker, CI/CD Pipelines (GitHub Actions, Jenkins)
Observability & Tools: Datadog, Dynatrace, New Relic, JFrog Artifactory, VS Code, Git/GitHub
Applied AI / Agentic Systems: LLM agents running in production (tool-calling, multi-agent skill design), agent orchestration, prompt & context engineering, MCP
AI-Augmented Engineering: Cursor, Claude Code, GitHub Copilot
Testing: Jest, Cypress, Enzyme, Detox, Mocha, JMeter, Fiddler
Methodologies: Event-Driven Architecture, Microservices, Clean Architecture, Domain-Driven Design (DDD), TDD, Agile/Scrum

Professional Experience

Makin

Tech Lead — Dec 2025 – Jul 2026

Led the technical delivery of a fintech onboarding platform, where security and anti-fraud were first-class concerns.

Drove technical strategy and architecture on NestJS and Next.js, integrating established market SDKs and vendors for identity, security, and anti-fraud.
Designed a robust Azure cloud architecture provisioned with Terraform, applying a design system to keep the product’s UI elements consistent.
Operated with a lean team and extensive use of AI agents to sustain delivery velocity without lowering engineering quality.

F3 Capital

Tech Lead & Senior Backend Specialist — Mar 2022 – Oct 2025

Acted as technical lead at a pension fund management company, driving the technical evolution of a strategic, highly regulated PrevTech SaaS platform designed to transform private pension operations.

Multi-Squad Coordination: Coordinated technical alignment and architectural consistency across multiple engineering squads, ensuring seamless system integration, unblocking technical dependencies, and scaling code quality standards for a large development team.
Core Product Delivery (Navega ERP): Contributed to the architecture and construction of a robust, full-scale Private Pension SaaS (Navega), successfully integrating an advanced, customizable “Engine of Rules” to execute complex native business logic securely within the pension fund environment.
Cloud & Serverless Infrastructure: Participated in the design and deployment of a resilient backend backbone using AWS Serverless (Lambda, SQS, SNS), achieving high horizontal scalability, strong security compliance, and cost efficiency for event-driven transactional workflows.
Engineering Excellence: Implemented a monorepo architecture and modularized internal shared libraries distributed via JFrog Artifactory, accelerating cross-squad feature delivery times and mitigating systemic code duplication.

Julius Baer Family Office

Fullstack Engineer — Apr 2021 – Apr 2022

Unified the digital user experience of two acquired wealth management companies (GPS and Reliance).
Built a seamless web and mobile application utilizing React Native Web tied to a centralized REST API Gateway written in TypeScript/Node.js.

Qintess

Fullstack Engineer — Apr 2020 – Nov 2021

Natura (Global Sales Platform - Technical Evolutions Squad): Embedded within a high-traffic squad tasked with increasing cloud application performance, scalability, and resilience.

Optimized microservices and serverless infrastructure, successfully boosting the system’s average Apdex from 0.79 to 0.96 within six months while reducing overall AWS architecture costs by 5%.

Devires Tecnologia

Fullstack Engineer — Aug 2019 – Mar 2020

QSaúde Ecommerce Portal & Mobile App: Engineered web and mobile solutions for health insurance onboarding and remote digital care (telemedicine, appointment scheduling, personal health questionnaires) utilizing Clean Architecture, React Native, Node.js, GraphQL, and Azure.

Amodev

Fullstack Engineer — Jun 2018 – Jun 2021

Budbuds.us: Developed a cross-platform agricultural control and evaluation mobile application built on React Native, GraphQL, Node.js, and PostgreSQL.
DONG (De Olho na Gôndola): Built a trade marketing application incorporating real-time GPS geolocation, dynamic backoffice data collection, and visit histories.
Mercado Bitcoin / Simulador Clube Bitcoin: Served as Front-end Architect to develop a cryptocurrency marketplace, building a full-scale market simulator leveraging ReactJS, Redux-Saga, and Cypress.

FCamara Training and Consulting

Fullstack Developer — Dec 2016 – Jun 2017

Developed core components for the Linkapi integration platform using Node.js, MongoDB, RabbitMQ, and TDD practices.
Formulated performance analytics for ASP.NET e-commerce sites using New Relic and Dynatrace.
Won the Big Hackaton-ONU at Campus Party 10 with the “Health Keeper” application (AngularJS/Node.js/MongoDB).

Applied AI — Personal Project

Operational AI platform for a small retail business — 2026 – Present

Designed and built a production system where an LLM agent runs as a live operator, not a coding assistant: a WhatsApp-based agent that executes real business workflows (inventory checks, cash-closing routines, staff task tracking, supplier ordering), backed by a custom agent framework (skills, ledger, middleware) and integrations with an ERP (Odoo) and a delivery marketplace (iFood).

Architected a skill-based agent runtime — each business workflow is a composable, testable skill rather than a monolithic prompt, with clear boundaries between deterministic domain logic and LLM-driven decisions.
Built connectors to external SaaS (ERP, delivery platform, messaging) so the agent acts on live operational data instead of static context.
Defined AI governance rules (revenue-first prioritization, human-in-the-loop gates for pricing and cash operations) so the agent supports the business without unsupervised financial actions.

Education

B.S. in Systems Analysis and Development — Fatec Carapicuíba (2013 – 2016)
Technical Degree in Web Design — ITB Maria Sylvia Chaluppe Mello (2012)

Languages

Portuguese: Native
English: Intermediate