Eduardo Cerqueira

Experience

Full career history

Featured projects

• Aletyx platform & AI gateway (Products & Platforms, 2025): Multi-cloud B2B SaaS with AI gateway, hardened containers, and a customer portal for tickets, artifacts, and release communications.
• Lyx for VS Code (Developer tools, 2026): I created and maintain this VS Code extension for the Aletyx engineering team. It brings the whole delivery flow into the editor: PRs and review metrics against a 4-hour SLA, each change tracked from PR to production, deploys and releases across GKE, ECS, and Cloudflare, Jenkins jobs, and an AI assistant that proposes actions you confirm. Its Agents tab and Ask Lyx run the @lyx agents from the editor, and Lyx also works as an MCP server in VS Code agent mode.
• Aletyx agents (@lyx) (AI & Bots, 2026): I built and run the AI agent platform behind @lyx for the Aletyx engineering team. Its agents scan and triage CVEs, explain failed Jenkins builds, run CI builds and tests, and manage sprint-board issues, from Slack or from VS Code. Built with LangGraph on Kubernetes; every run is traced in Langfuse and every action is confirmed by a person.
• Apache KIE (Open Source & Packaging, 2024): Open-source business automation community (KIE). Contributor and committer work alongside IBM Business Automation Manager Open Editions leadership.
• Jenkins Live Wall (DevOps & Cloud, 2026): I created and maintain this official Jenkins plugin under the jenkinsci organization. It turns jobs into a wall of big coloured tiles for real-time pipeline monitoring — built for the television in the corner of the office, so teams get instant visibility from across the room.
• Jenkins Concierge (AI & Bots, 2026): A Jenkins plugin I wrote and submitted to the Jenkins project: onboarding tours and an AI assistant that know your instance. Everything is computed from each person's own permissions, logs are stripped of secrets before they reach the model, and the assistant only proposes builds; nothing runs until you confirm.
• Fedora Copr packages (Open Source & Packaging, 2024): Maintains Fedora-related builds via Copr — including cloud-init, OpenStack tooling, and related packages — continuing a long open-source packaging practice from Red Hat years.
• Red Hat Insights / Hybrid Cloud Console QE (Products & Platforms, 2021): Years of Principal-level engineering on Insights SaaS quality: Kafka-backed microservices, CDN, automation frameworks, and CI/CD for console.redhat.com.
• AI Alpha Squad (AI & Bots, 2026): Autonomous multi-agent software delivery squad: open a GitHub Issue and specialized agents analyze, design, code, test, review, and open a PR — with a human Director approving key milestones.
• getalight.me (Products & Platforms, 2026): Free, ad-free screen lighting for video calls and recordings — open a spare monitor, tune brightness and Kelvin color temperature, then go fullscreen. Ships as a Cloudflare Workers web app plus a Chrome MV3 extension for preferred-monitor placement.
• AI Sandbox (AI & Bots, 2026): Personal monorepo of small, runnable AI experiments — agents, RAG, Hugging Face, LangChain, CrewAI, and MCP — each app self-contained and easy to tear down.

Expertise

Software Engineering

Architecture and hands-on delivery across distributed systems, APIs, and full-stack product work.

• Distributed systems & Kafka-backed services
• Python, Java/Spring Boot, TypeScript/Node
• REST APIs, SSO (Keycloak / Entra), multi-database designs
• Testing strategy: unit, integration, E2E, performance
• IDE tooling: VS Code & IntelliJ extensions, marketplace publishing

AI Agents & Automation

Agents in production for engineering and DevSecOps — built on open protocols, traced end to end, and confirming with a person before they act.

• Agent development with LangGraph and LangChain on Anthropic and OpenAI models, in containers on Kubernetes
• Protocols and SDKs: MCP servers and tools, A2A, Claude and OpenAI SDKs, VS Code agent mode
• LLM and agent observability: Langfuse traces, Grafana, Redis Insight, and an audit trail of every agent decision
• Guardrails: allow-listed actions, secret redaction, prompt-injection detection, human confirmation
• LLM gateways and RAG pipelines with token metering

Engineering Productivity with Agents

Agents wired into the team’s daily flow, from issue to production, so engineers spend their time on code instead of chasing status.

• Issue management and triage: agents open, update and file issues on the sprint board from Slack and VS Code
• CVE scanning, triage and fix proposals, with a rationale for every decision and nothing applied without review
• Code review with AI agent scores beside human reviewers, a 4-hour review SLA and merge times under one day
• Pipeline integration: failed builds explained with root cause and suggested fix; CI builds and tests run on request
• Release distribution: a guided PR-to-production checklist, signed artifacts and “CVEs fixed since the last release” reports

Cloud & Infrastructure

Multi-cloud platforms, hardened delivery, and CI/CD built for regulated and high-trust environments.

• Cloudflare Workers, Zero Trust, AI gateway
• AWS, GCP, Azure; Kubernetes / OpenShift
• Terraform, Ansible, GitHub Actions, Jenkins
• CVE pipelines; SOC 2 audit-ready, FedRAMP and FIPS delivery
• Signed artifacts, promote-by-digest, hardened images

Open Source & Leadership

Open source as a long-running practice — packaging, community, and leadership — not a portfolio costume.

• Jenkins Live Wall creator & maintainer (jenkinsci)
• Concierge: AI assistant & onboarding plugin for Jenkins
• Apache KIE committer / contributor
• Fedora Copr packaging (cloud-init, OpenStack tooling)
• Led teams of 8 to 60+ engineers, directly and through engineering managers
• Hiring, mentoring, and scaling remote engineering orgs