- LLM in production: a two-agent pipeline (OpenAI via LangChain4j): one simulates a senior recruiter (score, strengths, issues by severity) and the other generates structured improvements per section; 1,000+ analyses in the first week with 100% uptime.
- Stack and delivery: Spring Boot 3.5 (Java 25), Apache Tika and Next.js/React; Docker, CI/CD with automatic deploy to staging and a manual gate to production (git flow); open source (AGPLv3, open-core model).
Luanderson Pimenta Mendes
AI Engineer • AI in the SDLC • DevOps & Developer Experience
Lisbon, Portugalcontato@luanderson.dev.br+55 67 99195-0581linkedin.com/in/luanderson-devgithub.com/Luanderson-Dev
Summary
Software Engineer with 4 years of experience in backend, DevOps and applied AI. My focus is bringing AI into the software development lifecycle: reviewing team workflows, identifying where AI helps at each stage of the SDLC and integrating it safely into daily work.
I created Agent Software Factory (open source, v0.1.0), which organizes AI agents into roles across the entire SDLC, with human decision gates, security guardrails and verification through tests and CI. I am building cc-observatory to measure sessions, blocks, tokens and cost of AI agents in development.
I bring experience in DevOps (CI/CD, Docker, Kubernetes, Azure and observability, with a 99.9% SLA and 60% lower MTTR), Agile (Scrum and Kanban with product and QA) and team enablement, through DevOps mentoring and a technical community. I use AI in my daily workflow (Claude Code, Codex). Postgraduate studies in progress in Applied AI Engineering and in AI and Machine Learning. Native Portuguese speaker, advanced English (C1).
How I Work
- Workflow before tools: I map each SDLC stage (discovery, planning, coding, testing, review and release) and only then decide where AI helps, with clear roles, inputs and outputs.
- Humans in control: decision gates, security guardrails and evidence (tests and CI on the reviewed commit) for every AI-assisted delivery.
- Cross-functional collaboration: daily work with engineering, product, QA and DevOps; in the factory, architecture, security, testing and platform each have their own role and review point.
- Team enablement: I run “Do Zero ao Deploy”, a DevOps mentorship program, and the AceleraDev community, helping developers adopt cloud, DevOps and engineering best practices.
Featured Projects
- AI across the SDLC: an orchestrator and specialized roles (PO, architect, security, DevOps, planner and QA) take an idea from discovery and brief to architecture, plan, parallel implementation, review, merge and release.
- Humans decide, agents deliver: approval gates at brief, plan and release; guardrails that keep agents away from secrets, deploys and infrastructure; every delivery is verified against tests and CI.
- Tool-agnostic: Markdown protocol with per-harness adapters (Orca validated; Claude Code, Codex and Gemini CLI on the roadmap), 13 ADRs, threat model, schema validated in CI and documentation in English and Portuguese; v0.1.0 released as open source (MIT), built by the factory itself (dogfooding).
- Measuring day-to-day use of AI agents: a project to give visibility into Claude Code sessions, subagents, blocks (sessions waiting on a human decision), tokens and cost.
- Architecture: native telemetry via OpenTelemetry to Prometheus, Loki and Grafana, plus hooks for real-time state; privacy by default (metadata only, personal attributes removed, everything local).
- Foundation released (v0.1 on PyPI): 12 ADRs, telemetry and hook contract validated against real sessions, CI on Linux, macOS and Windows, strict mypy and a test coverage gate.
Experience
- Build and orchestrate AI agents in Python for decision support, replacing repetitive manual work in real operations.
- LLM integration into platform workflows through OpenRouter (multi-model gateway) and local models with Ollama, selecting each model by technical fit and cost per use case.
- AI-assisted development every day with Claude Code and Codex (implementation, refactoring and code review), with human review and test verification.
- Backend and data on PostgreSQL; standardized environments with Docker; React and TypeScript interfaces, working with distributed product, frontend and DevOps teams.
- AI automation: chatbots and automations in Python and n8n integrated with corporate platforms, reducing manual work in support and operations.
- Observability and reliability: Prometheus, Grafana, Loki and Zabbix stack; led incident investigation and root cause analysis, with a 99.9% SLA and 60% lower MTTR.
- Cloud and containers: Microsoft Azure, Docker and Kubernetes, delivering a 40% gain in operational efficiency; CI/CD pipelines with GitHub Actions, Azure DevOps and Jenkins.
- Networking: MikroTik and UniFi, VLANs, firewall and VPN.
- Servers and data: Windows and Linux servers, Active Directory, and database performance tuning (PostgreSQL, SQL Server, MongoDB and Redis).
- Agile teams: Scrum and Kanban, with technical documentation of environments and processes.
- Own AI products: Evolui CV (LLM in production, 1,000+ analyses in the first week), Agent Software Factory and cc-observatory, described under Featured Projects.
- End-to-end delivery, from requirements gathering with stakeholders to deploy: backend in Java (Spring Boot), Node.js, TypeScript, Python and C# (ASP.NET Core); frontend in React, Next.js and Angular; Docker, CI/CD and automated tests (JUnit/Mockito, xUnit/Moq).
Education
- Topics: AI applied to DevOps and software development, autonomous agents and Agentic AI, Model Context Protocol (MCP), RAG, AI-first architectures, AI security and governance.
Skills
AI in development: AI agents and multi-agent orchestration, LLMs (OpenAI, OpenRouter, Ollama), LangChain4j, Claude Code, Codex, prompt engineering, n8n; MCP and RAG (in training)
SDLC and practices: Agile (Scrum, Kanban), TDD, code review, ADRs, threat modeling, Git and GitHub flow
DevOps and platform: CI/CD (GitHub Actions, Azure DevOps, Jenkins), Docker, Kubernetes, Azure, AWS, Prometheus, Grafana, Loki, OpenTelemetry
Engineering: Java (Spring Boot), Python, Node.js, TypeScript, C# (.NET), REST APIs, JWT, microservices, PostgreSQL, SQL Server, MongoDB, Redis, Kafka, RabbitMQ, Elasticsearch
Languages: Portuguese (native), English (advanced, C1), Spanish (intermediate, B1), Italian (intermediate, B1)