Miguel Oliveira

Engineering Manager · Developer productivity, AI-SDLC & agentic test automation

I drive the shift from manual QA to company-wide automation at Lively, a B2B SaaS company. Together, the team took regression from a three-day manual cycle staffed by seven engineers to about twenty minutes in CI, and moved the organisation from a release every two weeks to two a week, with far fewer hotfixes. My part, from the foundation of that transition, has been the direction: the strategy and what to build, growing manual testers into automation engineers, and the architecture of the end-to-end framework we built from the ground up. Most of our work now sits at the intersection of agentic systems and engineering velocity: embedding LLMs into the CI pipeline, and an "AI-led, human-validated" delivery model adopted across the engineering organisation.

The throughline across thirteen years is the same: verification at every layer of the stack, from RTL to LLM agents. Based in Aveiro, Portugal.

Currently shipping

• Replaced an unmaintained ~100-case WebdriverIO suite with an end-to-end Playwright and TypeScript framework the team built from scratch: independent tests, Allure reporting, and fully sharded parallel runs on Microsoft Playwright Service.
• Every pull request now validates against a dockerized build in CI, with the full ~20-minute regression gating each release: the gate that lets the org release twice a week instead of every two weeks.
• Extended the platform sideways with dedicated frameworks for load and performance testing (k6) and for LLM validation (DeepEval with Python).
• Agentic test-autofix on the Anthropic SDK and MCP, and the "AI-led, human-validated" delivery model I introduced, now adopted across the engineering organisation.

Career

Thirteen years of verification and engineering-productivity work, moving up the stack one layer at a time, from silicon to software to AI.

• 2023 –

  Automation Engineering Manager · Lively Joined as Senior QA Engineer and moved into engineering management. Drive the company-wide move from manual QA to automation, setting the strategy and growing the team that built an end-to-end framework from scratch, taking regression from three manual days to about twenty minutes in CI and doubling release cadence. Established the "AI-led, human-validated" delivery model adopted across engineering.

• 2022

  Senior Automation Engineer · Urbint May–Dec 2022. Ran a criteria-based evaluation of the existing Cypress suite against a Playwright proof-of-concept (maintainability, parallelism, test-data handling) and, on its outcome, built a TypeScript Playwright framework from scratch with first-class test-data management and reporting at its foundation.

• 2022

  Team Lead QA Engineer · 8x8 Feb–May 2022. QA leadership for a UCaaS communications platform.

• 2019 – 22

  QA Engineer → Team Lead QA Engineer · Fuze Led a 5-person QA team on a UCaaS desktop and web product. Led the project to replace the native Fuze desktop client with an Electron app built on the existing web implementation, and built Selenium/Java automation (Appium, WinAppDriver) on Jenkins pipelines.

• 2013 – 19

  Verification Engineer · Synopsys Porto ASIC digital verification on DDR PHY and MIPI DPHY IPs: SystemVerilog/UVM testbenches, Verilog modelling, and post-layout simulation.

• 2013

  MSc, Electrical & Computer Engineering FEUP, University of Porto.

Selected skills

• Verification & quality: SystemVerilog/UVM hardware verification; Playwright, Cypress, Selenium, k6 and DeepEval across web, desktop, load and LLM testing; quality strategy owned at the team and platform level.
• AI-SDLC & agentic systems: Anthropic SDK and MCP; LLM evaluators and agentic autofix in CI; the "AI-led, human-validated" delivery model.
• Developer productivity & CI/CD: GitHub Actions, Jenkins, Codefresh; GCP; Datadog and Splunk; TypeScript, JavaScript and Java.
• Leadership: getting a team to pull in one direction and carrying initiatives from first idea to shipped; building and leading quality and automation teams; onboarding and mentoring engineers; OKRs and Management 3.0.

Open to

• Technical conversations on AI-SDLC, agentic testing, and platform quality engineering
• OSS collaboration in the space

Contact

Colophon

This page is one self-contained HTML file: inline CSS, no webfonts, no JavaScript framework, no analytics, no third-party requests beyond a crawler-only Open Graph image. It respects dark mode and reduced motion, targets WCAG AA contrast, and prints to a clean one-page CV. Same bar I hold engineering teams to. View the source.