Skip to content

I use Google Analytics to understand what gets read, shared and clicked here. It only loads if you accept — nothing third-party runs before that. How Google uses this data

Luiz Tadeu Mendonça

AI Engineer — Context & Harness Engineering | Claude Code · Kiro · Python · Node.js · TypeScript · AWS · Terraform | 18+ years across Software Development & Distributed Systems · São Paulo — Brazil

AI Engineering applied to agile software development — Claude Code, Kiro, AI-DLC and Context & Harness Engineering — to design, build and ship production-ready systems. That's the newest layer on 18+ years of software engineering, not a fresh start: I began in enterprise integration and packaged software, moved into modern product engineering shipping web and native-mobile apps, and have spent the last years building modern applications on AWS — solution architecture, application and infrastructure code, and the technical direction of the build — across regulated industries like financial services, aerospace and energy. My lane is applied GenAI and agentic development, not machine learning research: I build with agentic patterns — tool-calling, RAG, memory, evaluation loops, MCP — and bring the SDLC rigor that turns AI work into production software. Python for AI, agents and backend; Node.js and TypeScript for the full-stack and web layer; Spring Boot for microservices.

01

Experience

2023-04 – PresentSenior Delivery Consultant — App ModernizationAmazon Web Services — Professional Services

Application-modernization and new digital-platform launch programs — microservices, full-stack web, smart TVs and native mobile. Leading the implementation: solution architecture, AWS infrastructure and the technical direction of the build. The launches: a custom cloud-native replacement for a SaaS streaming platform, native across five platforms plus Web, and — as tech lead — an upstream operational-monitoring platform on an oil & gas operator's AWS landing zone. Embedded in enterprise engineering teams delivering cloud-native, distributed systems end-to-end — from application code to infrastructure as code — while moving delivery into an AI-native loop with Claude Code, Kiro, AI-DLC and Context & Harness Engineering practices.

  • Adopted AI-native development (Claude Code, Kiro, AI-DLC and Context & Harness Engineering) in 2026 to design, build and ship production-ready systems, and am structuring the practice for team-scale adoption.
  • Architected the internalization of a third-party SaaS streaming platform and the modernization of that workload on AWS — native apps across five platforms (iOS, Android, Tizen, webOS, Apple TV) plus Web, on a BFF + microservices backend, designed for future B2B enablement — from envisioning and team setup through to steering multi-vendor delivery.
  • As tech lead, stood up an oil & gas operator's AWS landing zone and delivered its first modernized workload — an upstream operational-monitoring platform — owning the infrastructure and coordinating a separate consultancy's development across a ~2-year engagement the client renewed on the strength of the deliveries.
  • Built, hands-on, the serverless data integration for an aerospace manufacturer after its Heroku-to-AWS migration — Salesforce into a modernized custom backend, on Terraform, Amazon AppFlow and Python.
  • Developing an internal full-stack web platform — implemented in AI-DLC, with a reusable Kiro harness customization other projects can adopt; embeddings-based semantic search running on Amazon Bedrock over Amazon S3 Vectors; integrated with inbound corporate Slack and with AI tooling over a bidirectional MCP server; MVP in progress — for sharing how AI is used in LATAM app-modernization delivery.
  • Lead engagements across LATAM, developing architectures, applications and delivery strategy, and presenting them to C-level and technical stakeholders in English, Spanish and Portuguese.
A man standing in front of a plain wall reading "aws Summit São Paulo", a lanyard and badge around his neck.
The event came back after the pandemic. I came back with it — for the first time as an employee.
2021-01 – 2023-03Cloud Application ArchitectAmazon Web Services — Professional Services

Application-modernization programs — full-stack web on a cloud-native stack. Hands-on individual contributor into tech lead: I set the platform up and wrote application and infrastructure code. Architected and implemented cloud-native solutions embedded within enterprise client teams, operating under legacy, compliance and data-sovereignty constraints.

  • Stood up the cloud-native foundation of a tier-1 bank's mortgage-credit domain, migrating it off mainframe and ASPX to a full-stack modern platform — Angular; Kotlin (Micronaut, later Spring Boot); EKS, ArgoCD and Istio; RDS, DynamoDB, SQS and SNS.
  • Grew from hands-on individual contributor to tech lead of a cross-organization team of up to eight over ~2 years; the open-source foundation kept expanding after my rolloff.
  • Built reusable Terraform modules and deployment patterns other engineering teams relied on, applying software-engineering rigor to infrastructure as code.
A man standing in the middle of a long, empty office corridor, circular light fittings overhead and the ceiling opened up to show ducts and pipework; lift doors along the right.
A technology I admired and wanted to use every day. Here the ceiling was open.
2020-06 – 2021-01Senior DevOps EngineerGlobo.com

New-platform launch program — a direct-to-consumer sales and subscription platform. Function: observability and DevOps engineering, instrumenting the web revenue path end to end. Owned observability for the sales / subscription journey of a major direct-to-consumer streaming service and its channels — a business-critical revenue path on Angular, Salesforce and Spring Boot.

  • Built an end-to-end observability platform integrating AppDynamics, Grafana, Prometheus and Zabbix into a unified monitoring solution.
  • Instrumented the Angular frontend and the Salesforce / Spring Boot backends so the revenue path had first-class visibility across the stack.
A smiling man behind the open lid of a laptop covered in tool stickers — Amazon Web Services, Elastic Stack, Terraform, Flutter, npm, VS Code, SonarQube, Docker, Kubernetes, MongoDB, Redis, Python, Android — wearing a t-shirt of a whale carrying containers.
My favourite stacks at the time, stuck to the lid.
2015-01 – 2020-06Digital Business Integration ConsultantAccenture

New-platform launch projects — full-stack web and native mobile — from 2017 at Accenture Digital; enterprise integration architecture before that. Application architect, hands-on across every tier: mobile clients, web front ends, the backends under them and the delivery pipeline — four custom-build engagements across four sectors. Promoted to Consultant, first as an integration architect on enterprise CRM implementation programs, then — the turning point — as an application architect at Accenture Digital, building modern web and native-mobile products. This is where the modern-engineering identity was forged and the through-line I carry today began — designing and building modern applications, and leading the build.

  • Architected the integration layers for enterprise CRM implementation programs, connecting apps, APIs and heterogeneous enterprise platforms.
  • Responsible application architect on a telecom field-force programme — three apps built in parallel, one hybrid (Ionic) and two native Android, the front technology fixed per app at requirements time by each user population's devices. Under fronts that were required to differ, the backends were much alike in architectural concept, because we traded what worked across the parallel teams. Hands-on at every tier: the Android clients, a React back-office portal, and the Node.js / Express services under them on MongoDB and Redis, offline-first throughout.
  • Designed the API contract and the non-relational data model so that synchronising needed no additional backend component at all — MongoDB on the server, SQLite on the device. Two off-the-shelf sync stacks had been tried on the hybrid app, RabbitMQ then CouchDB, and neither held for this workload, so we moved offline-first out of the infrastructure and into the design.
  • Ran the full delivery on both native Android apps — the field technician's, then the field salesperson's, where the model proved out: from look-and-feel ideation with the UX team, through the event trail that made them operable in the field, to the analytics that stakeholders read.
  • Collaborated on the delivery of a Flutter app covering iOS and Android for a digital bank; built the DevOps and delivery pipeline for a commerce platform; and, in the final Accenture years, guided the reversal of a recurring performance problem in a Java / Spring Boot microservices backend running on AWS, for an education product.
  • Left packaged software behind for custom, lean-stack product engineering — open source, private-cloud infrastructure, and hands-on distributed-systems design. 2017 is the seam: from there on, the work has been platforms being built rather than packages being implemented.
  • Grew into the through-line I carry today — designing and building modern applications and leading the build on delivery teams, as a senior individual contributor.
A man in glasses in the foreground, turned to the side, with a desk behind him: an external monitor showing a monitoring dashboard of charts, and an open laptop showing a dark-theme code editor.
The last months of this one were at home, with nobody around.
2008-03 – 2015-01Systems Integration AnalystAccenture

Enterprise-integration SDLC — packaged implementations, ETL and SOA. Hands-on build and integration of large-scale distributed systems. The root of the career: an enterprise integration factory. Began in web development, then built and integrated large-scale distributed systems — batch (Informatica PowerCenter ETL) and online / real-time (SOA) — growing from junior to senior analyst.

  • Joined as an intern in 2008 while completing the Information Systems degree; from graduating in 2010, grew from junior to senior analyst building enterprise system integrations.
  • Built and connected large-scale distributed systems across batch (Informatica PowerCenter ETL) and online / real-time (SOA) integration, wiring heterogeneous enterprise platforms together.
  • Delivered across the full SDLC for enterprise clients in regulated sectors — financial services, telecom, e-commerce.
A group of people in front of an office building reading "accenture": five standing across the front — the one at the left edge only half in frame — and two more behind them, smiling at the camera; some with badges on lanyards around their necks and tall towers closing the background.
Manila. This job took me to the other side of the world.
02

Education

2006 – 2010Bachelor's Degree, Information Systems — electives in Databases and Operating SystemsPontifícia Universidade Católica do Rio de Janeiro (PUC-Rio)
03

Certifications

AWS Certified Solutions Architect – ProfessionalAmazon Web Services (AWS)
AWS Certified AI PractitionerAmazon Web Services (AWS)
AI-DLC AmbassadorAmazon Web Services (AWS)
AWS Accreditation — AI (L200)Amazon Web Services (AWS)
AWS Accreditation — AI (L100)Amazon Web Services (AWS)
AWS Certified Developer – AssociateAmazon Web Services (AWS)
AWS Certified Solutions Architect – AssociateAmazon Web Services (AWS)
AWS Well-Architected ProficientAmazon Web Services (AWS)
AWS Accreditation — Telecom (L100)Amazon Web Services (AWS)
04

Skills

AWS Cloud
AWS LambdaAmazon API GatewayAmazon S3Amazon DynamoDBAmazon ECSAmazon EKSAmazon SQSAmazon SNSAWS WAFAmazon CognitoAWS KMSAmazon RDSAmazon CloudFront
Distributed Systems & DevOps
Distributed Systems ArchitectureMicroservicesServerless ComputingEvent-Driven ArchitectureTerraformAWS CloudFormationCI/CDSecurityPlatform Engineering
Observability
Amazon CloudWatchAWS X-RayPrometheusGrafanaELK StackSplunk
AI-native Engineering
AI-DLCContext & Harness EngineeringClaude CodeKiroContext EngineeringAgents EngineeringMCP (Model Context Protocol)
AI Engineering
Large Language Models (LLM)Prompt EngineeringRAG (Retrieval-Augmented Generation)Amazon Bedrock
Languages
PythonTypeScriptNode.jsJavaSpring Boot
Languages (spoken)
PortugueseEnglishSpanish