Walber Fellipe— Fullstack EngineerAI Focused
6+ years building products end-to-end with emphasis on AI engineering: LLM integrations, agentic workflows, RAG and evaluation, and robust APIs—combined with React, Next.js, and cloud platforms to ship reliable experiences for real users.
6+
Years of mastery
Market impact
20k+
Users impacted through Cibus platform optimization and core systems.
Scale & performance
15k+
Concurrent users managed for Strings Brasil infrastructure.
Leadership
Active
Leading high-performance engineering teams at Diagonal architectural group.
Meta · this site
Shipped with an AI-native loop—not a template story
The codebase was refactored end-to-end: layouts, project catalog, experience page, and case-study shells now follow a dedicated Figma spec instead of carrying forward an old starter-site narrative.
Cursor (agent-style sessions and Composer) handled most of the implementation work. For Experience, retrieval over my own CV (RAG) kept dates, roles, and stack aligned with what I actually shipped. The Figma MCP let the assistant read frames directly so UI work matched design instead of guesswork.
Google Stitch generated first-pass page ideas; those went into Figma for layout and typography polish, then back into Next.js and Tailwind here. The case study for this site spells out the same pipeline in more detail.
Read the My Portfolio case studyCore Technology Stack
A strategic selection of tools for scalable products—including AI-ready APIs, typed clients, and observable backends.
Stack selection 2026
AI Engineering & Agentic Workflows
Orchestrating LLMs with RAG and MCP to build autonomous agents. I focus on systems that don't just "chat," but execute tasks, call tools, and use proprietary data with clear boundaries.
I'm currently enrolled in Software Engineering with Applied AI (second module): generative AI APIs, provider landscapes (OpenAI, Anthropic, Hugging Face, Gemini), advanced prompt engineering (chaining, templates, reducing hallucinations), cost and token discipline, deeper RAG with orchestration (e.g. LangChain-style flows), observability and debugging in AI apps, plugging LLMs into real backends, and an intro to multimodal models—building on earlier modules in ML/DL on the web, embeddings, and MCP.
Production-grade prompts, evals, and guardrails.
AWS
S3 for durable object storage, assets, and safe document handoffs; Cognito for hosted auth, tokens, and user pools integrated with APIs and SPAs.
React & Next.js
Immersive UX with SSR, edge functions, and patterns that keep complex dashboards—including AI copilots—fast and maintainable.
Web Vitals optimized
Node.js & NestJS
Backend services built on NestJS (modules, DI, guards) on top of the Node.js runtime—structured APIs, clear boundaries, and patterns that scale with the team.
ORMs & databases
From Prisma and Drizzle schemas to production Postgres, MongoDB, MySQL/MariaDB—migrations, queries, indexing, and data modeling across the stack (and whatever the next project throws in).
Case Study: Regulariza.ai & Titula.ai
Technical feat
Architected data verification flows with serverless processing and dependable real-time state for document-heavy journeys.
Solving real-world compliance
Engineered core flows for Titula.ai and Regulariza.ai around automated document processing and clear compliance paths on the modern web stack.
- Automated verification pipelines
- Next.js edge function optimization
- Secure AWS infrastructure management
Let's build something extraordinary.
Open to architectural consulting, AI product delivery, and high-impact full-stack collaborations.