Architecture That Scales.
AI That Ships.
Engineering That Delivers.

Senior Full-Stack Engineer & AI Architect. I help companies design distributed systems, build production AI features, and ship products that compound — and I mentor developers who want to level up their careers in architecture and AI. 6+ years at enterprise scale.

Trusted by teams at

SulAméricaZurich SantanderKoruResilionSOL CopérnicoPowerOneEximiaTI

Architecture + AI.
One Senior Engineer.
No Agency Overhead.

I work across the full stack — from system architecture to AI features in production. Whether it's designing a backend, building an AI copilot, or helping a team ship faster, I focus on solving the actual problem with the simplest approach that works.

6+
Years Exp
50+
Systems
5+
AI Systems
2
Products
20%
AWS Reduction
6
ML Models
Jônatas Fernandes Pimenta

The Short Version

I started coding at 12 with RPG Maker and building neural networks by 17. That curiosity turned into a career spanning enterprise architecture, AI engineering, and product development — building mass communication platforms for companies like SulAmérica and Caixa Seguradora, an IoT intelligence platform with AI copilot at Koru, and reducing a 12-month document workflow to 2 months at Zurich Santander.

Today I work with companies of all sizes — from startups to enterprises — who need someone to architect scalable systems, ship real features, and unblock their engineering teams. I also build my own products: Pocket Books (a reading platform with AI features), Cronock (a B2B2C booking SaaS), and ClaudIA (an open-source multi-agent assistant). The best engineers understand what it means to own something end-to-end.

MBA in Software Architecture (2025) · MBA in AI Engineering (2026) · Portuguese (native) · English (fluent)

What I Actually Do

neurology
Most Requested

AI Agents, RAG Pipelines & Intelligent Products

Production AI systems, not prototypes. LangGraph multi-agent architectures, RAG pipelines with vector search, AI copilots with tool-calling, recommendation engines, ML model ensembles, and structured LLM outputs.

LangGraphLangChainOpenAIRAGCopilotKitAI Agents

Architecture Consulting

Designing the right architecture for your problem — not the most complex one. Monoliths when they make sense, microservices when they don't, and the pragmatic middle ground most teams actually need.

account_tree

Product Engineering

Full-stack execution from Figma to production. Modern tech stacks built for speed and maintainability.

deployed_code

System Integration

Connecting fragmented legacy systems, universal hardware integration, and IoT ecosystems.

hub

Interim CTO

Strategic technical leadership, hiring roadmaps, and engineering culture for startups.

shield_person

Proven Results. Across Industries. At Scale.

A selection of enterprise projects and strategic technical interventions.

Omnichannel Communication Platform — 30+ Workflow Nodes, Git-Like Versioning
Enterprise SaaS · InsuranceCxMapper

Omnichannel Communication Platform — 30+ Workflow Nodes, Git-Like Versioning

Built a SaaS platform for visual communication automation used by SulAmérica and Caixa Seguradora. Visual DAG workflow builder with 30+ node types, unique Git-like versioning system (branches, commits, PRs, approval gates), WhatsApp chatbot orchestrator, 4 template types, multi-tenant architecture with RBAC, and full observability. NestJS monolith consolidated from 20+ microservices.

30+
Workflow Nodes
2
Enterprise Clients
40M+
Msgs/Month
NestJSNext.jsPrismaPostgreSQLAWS SESWhatsAppOpenTelemetry
Infrastructure Intelligence Platform — AI Copilot + IoT Sensor Network
AI · IoT · InfrastructureResilion / Koru

Infrastructure Intelligence Platform — AI Copilot + IoT Sensor Network

Key contributor to the Resilion platform, working across multiple systems: helped build the Nexus backend (Hono + PostgreSQL/TimescaleDB), Sentinel monitoring dashboard (Next.js 15 + Mapbox + Three.js), AI Copilot (LangGraph with 15+ tools and intent classification), Ranger data collector agent, and Courier IoT gateway (protocol-agnostic FTP/SFTP/MQTT). Contributed to the design of a 6-model ML ensemble for predictive failure forecasting.

5
Systems Built
15+
AI Tools
6
ML Models
HonoNext.js 15LangGraphCopilotKitTimescaleDBMQTT
Document Processing: 12 Months → 2 Months — 6x Acceleration
Financial Services · EnterpriseZurich Santander

Document Processing: 12 Months → 2 Months — 6x Acceleration

Acted as business analyst on the Zurich Santander retroactive reading initiative, reengineering the document-processing workflow to reduce a projected 12-month timeline to just 2 months through strategic pipeline parallelization and infrastructure tuning.

6x
Faster
80%
Automation
~€2M
Annual Savings
Pipeline ParallelizationInfrastructure TuningBusiness Analysis
Enterprise Integration Hub — 8 ADRs, 82 Business Rules, 8 External Systems
Insurance · Software ArchitectureZurich Santander

Enterprise Integration Hub — 8 ADRs, 82 Business Rules, 8 External Systems

Authored a complete Solution Architecture Document (SAD) with 8 ADRs, catalog of 82 business rules across 7 categories, 37 detailed use cases, and a complete data model with 4 schemas. Specified integrations with 8 external systems (Alfresco ECM, Jarvis OCR, Pega BPM, CCM, Service Provider, Corporate Portal, BOD Seguros, PostgreSQL) with resilience policies (Circuit Breaker, Retry, Rate Limiting). Defined 8 Bounded Contexts using DDD tactical patterns and planned a gradual migration using Strangler Fig Pattern with canary deployments and automatic rollback.

8
ADRs
82
Business Rules
8
Bounded Contexts
DDDStrangler FigSpringAlfresco ECMPega BPMResilience4jAzure
3 AI Pipelines — Recommendations, Retrospectives & Reading Plans
Consumer AI · SaaSPocket Books

3 AI Pipelines — Recommendations, Retrospectives & Reading Plans

Built three production AI pipelines: a two-phase LangGraph agent for reading plans (20-iteration tool-calling loops against Google Books API), a recommendation engine with recency-weighted context (GPT-4o, 7 candidates monthly), and a retrospective generator with 13 structured AI-generated text fields. All async via SQS, premium-gated, multi-language.

3
AI Systems
20
Agent Iterations
7
Candidate Engine
LangGraphOpenAI GPT-4oNestJSStripeSQS
Integration Hub + AWS Optimization — 60% Less Manual Work, 20% Cost Reduction
Solar Energy · EnterpriseSOL Copérnico

Integration Hub + AWS Optimization — 60% Less Manual Work, 20% Cost Reduction

Built an Integration Hub connecting sales platform orders with SAP, reducing manual work by 60% in the first phase. Implemented a new tech stack (NestJS, Next.js, React Native, Fargate). Structured the company's microservices architecture — defined when to use microservices, which services should stay monolithic, and migrated some microservices back to monoliths. Achieved a 20% cost reduction on AWS. Also built a subscription system, invoice app, automated bill extraction bot, and the Generator System for assembling solar panel kits.

60%
Manual Reduction
20%
AWS Cost Reduction
5+
Systems Delivered
NestJSNext.jsReact NativeAWS FargateSAPSalesforce
Custom Frame E-commerce — Full Product from Database to Deploy
E-commerce · StartupQuadroo

Custom Frame E-commerce — Full Product from Database to Deploy

Built the full Quadroo platform from scratch — an app where customers purchase custom frames with personalized photo styles. Handled database modeling, API development, payment gateway integration, Brazilian postal service (Correios) integration, AWS deployment, and the company's inventory management system.

E2E
Full Product
3+
Integrations
AWS
Infrastructure
Node.jsReactPostgreSQLAWSPayment GatewayCorreios API

Engineering Philosophy

Keep it simple. I don't pick architectures because they're trendy — I pick them because they solve the problem with the least complexity possible. A well-structured monolith beats a poorly designed microservices mess every time. DDD when the domain warrants it, event-driven when you actually need decoupling, and a straightforward API when that's all you need. The same pragmatism applies to AI: structured outputs, proper error handling, real data validation. The right solution matches your constraints, not the hype cycle.

// domain/order.service.ts (DDD)
@Injectable()
export class OrderService {
  async createOrder(cmd: CreateOrderCommand) {
    // Transactional consistency boundary
    return await this.uow.execute(async (ctx) => {
      const order = Order.create(cmd);
      await this.repo.save(order, ctx);
      this.events.dispatch(new OrderCreated(order));
    });
  }
}
# ai/agent_workflow.py (LangGraph)
workflow = StateGraph(AgentState)
workflow.add_node("reason", reasoner_node)
workflow.add_node("tool_call", tool_executor)

workflow.add_conditional_edges(
    "reason",
    should_continue,
    {"continue": "tool_call", "end": END}
)

app = workflow.compile()
For Individuals

Mentoring & Career Coaching

Not everything I do is for companies. I also work 1-on-1 with developers who want to grow faster — whether that's landing a senior role, transitioning into architecture, or breaking into AI engineering.

trending_up

Career Acceleration

Strategic guidance to get unstuck and reach the next level. We work on positioning, interview prep, salary negotiation, and building a career path that actually matches your goals.

Junior → SeniorInterview PrepSalary NegotiationCareer StrategyPortfolio Review
code

Technical Mentoring

Deep-dive sessions on system design, architecture patterns, and production-grade engineering. I help you think like a senior engineer — not just write code, but make decisions.

System DesignDDDMicroservicesClean ArchitectureCode Review
neurology

AI & ML for Engineers

Practical guidance for developers who want to add AI to their skillset. From building your first LangGraph agent to shipping RAG pipelines — real projects, not tutorials.

LangGraphRAG PipelinesLLM IntegrationAI AgentsML Fundamentals

Ready to Level Up?

Book a free intro call. We'll talk about where you are, where you want to go, and whether I can help.

Book a Free Intro Call

Products I've Built, Not Just Deployed

Pocket Books
AI-Powered Reading Platform

3 production AI pipelines: LangGraph reading plan agent, GPT-4o book recommendation engine, and retrospective generator (like Spotify Wrapped for readers). NestJS backend, Stripe payments.

Explore arrow_forward
ClaudIA
Multi-Agent AI Assistant

CLI productivity assistant with LangGraph multi-agent architecture (Main, Planning, Summary agents). RAG knowledge base via Qdrant vector database. Google Calendar + Microsoft Calendar integrations.

Explore arrow_forward
Cronock
B2B2C Appointment Booking SaaS

Appointment booking platform for beauty and aesthetic services in Brazil. NestJS + PostgreSQL with DDD/Clean Architecture, React Native/Expo mobile app, MercadoPago payments.

Explore arrow_forward
CmdMMO
Terminal-Based MMORPG

A multiplayer roguelike MMO that runs entirely in the terminal. Real-time multiplayer via Socket.IO, 3 classes, procedural dungeons, turn-based combat, crafting, housing editor, and global rankings. Built with a custom engine (GlyphEngine). 19 stars on GitHub, 57 upvotes on TabNews.

Explore arrow_forward
GlyphEngine
Terminal Game Engine

A reusable Python framework for building ASCII-based terminal games. Handles rendering, input, game loops, and UI components. Extracted from CmdMMO as a standalone engine for anyone to build terminal-based games.

Explore arrow_forward

I Share What I Learn

Published writing on architecture, distributed systems, and engineering leadership.

Apr 14, 2026

Pare de Colocar LLM em Tudo

Um guia sem hype — com casos reais, custos e o framework de 3 perguntas para saber se você realmente precisa de um.Existe muito hype em IA. Todo mundo quer colo…

product-managementllmai-agent
Sep 23, 2024

How to establish processes in your business

Systemizing the predictable, so you can humanize the exceptional — Isadore SharpProcess can be a scary word, mainly for those who like freedom. Well, what if I …

businessentrepreneurshipprocess-improvement
Sep 20, 2024

Como priorizar uma feature?

Ja se sentiu perdido na hora de priorizar uma tarefa, ou, até mesmo, alguma vez tomou alguma decisão que parecia a certa no momento e depois gerou mais prejuízo…

project-managerproject-managementprojetos
Feb 26, 2024

Conhecendo a Arquitetura Space-Based

Em qualquer aplicação de alto volume onde existe muitas transações concorrentes, o banco de dados normalmente vai ser o ultimo fator que irá limitar a quantidad…

backendarquitetura-de-softwaresoftware-architecture
Dec 4, 2023

Explorando a Arquitetura Baseada em Serviços: Flexibilidade Arquitetural em Foco

A arquitetura baseada em serviços, um híbrido da arquitetura de microsserviços, é amplamente reconhecida pela sua flexibilidade arquitetural. Embora seja uma ar…

software-architecturesoftware-engineering
Nov 29, 2023

Desvendando a Arquitetura Microkernel: Simplicidade e Flexibilidade em Foco

A arquitetura microkernel (também conhecida como plug-in) foi concebida há várias décadas e ainda é amplamente utilizada nos dias de hoje. Essa estrutura é idea…

pluginsbackendplugin-development
Nov 27, 2023

Desmistificando as Falácias dos Sistemas Distribuídos

Os sistemas distribuídos, especialmente em ambientes de microservices, apresentam uma série de desafios que muitas vezes são subestimados. Vamos explorar alguma…

software-architecturearquitetura-de-softwaresoftware
Nov 27, 2023

Desvendando as Métricas de Arquitetura: Abstração, Instabilidade e Distância da Sequência Principal

No vasto campo da engenharia de software, a arquitetura de um sistema desempenha um papel vital na determinação de sua eficiência, manutenibilidade e estabilida…

software-architecturearquitetura-de-software
Nov 27, 2023

Dominando o Uso de Keys no React

No vasto ecossistema do React, entender o papel das chaves (keys) é essencial para otimizar o desempenho e garantir uma renderização eficiente dos componentes. …

reactnextjsreactjs
Nov 27, 2023

Otimizando o Desempenho no React: Memo, useMemo e useCallback

No universo do desenvolvimento React, a eficiência é um fator crucial para proporcionar uma experiência de usuário suave. Ao compreendermos os casos em que um c…

reactjsreact-hookreact

Let's Talk

Whether you need an engineer for a high-impact project or a mentor to accelerate your career — I'd love to hear from you. Briefly describe what you're looking for and let's see if we're a fit.

Send Inquiry