Senior Director, Software Development, Test Automation

About this role

Your Impact at LILA

The Role

We're hiring a Senior Director, Software Development, Test Automation Systems to architect and build Lila's test automation platform and quality engineering practice for our AI-powered scientific and lab automation products. Reporting to the VP of Engineering, you'll own the test automation system, CI/CD test infrastructure, AI-driven test tooling, and the eval discipline that hold the bar across our SDLC.

This is a builder-leader role. You will drive the quality vision, write requirements, make sharp build-vs-buy calls, drive execution, and build and lead a small (3–5 person) team that delivers leverage. The operating model is federated: you own the platform, standards, and metrics; engineering teams own test execution. You scale through tooling and influence.

As you scale into this role, you'll also stand up the QC framework for our lab automation system — the validation patterns, harnesses, and contracts that science operations teams will operate day-to-day. Data integrity and ALCOA+ compliance are foundational to everything you build.

What You'll Be Building

What You'll Do

Architect and ship the test automation platform

• Design and build the test automation platform — frameworks, fixtures, golden datasets, test orchestration, and reporting — that the engineering org adopts by default

• Set standards across unit, integration, contract, end-to-end, regression, performance, and chaos testing for backend services, the frontend monorepo, and data pipelines

• Treat platform adoption, flake rate, and time-to-signal as first-class engineering metrics

Make build-vs-buy decisions with conviction

• Own the buy/build/borrow strategy across test infrastructure, eval platforms, browser/device clouds, observability, and lab QC tooling

• Justify every choice with TCO, signal quality, integration cost, and time-to-leverage — and revisit decisions as the org and tech landscape evolve

• Bias toward leverage: buy commodity capabilities, build the differentiators (Lila-specific AI evals, lab QC, scientific data integrity)

Modernize CI/CD for fast, reliable signal

• Own the test execution layer of CI/CD: parallelization, caching, hermetic environments, ephemeral preview envs, and affected-only test selection across our Nx monorepo/microservices.

• Build retry, quarantine, and impact-analysis systems so signal stays sharp as the org scales

• Drive change-failure rate, MTTR, Test effectiveness, pipeline efficiency, coverage, and PR-to-prod lead time as outcomes

Drive AI-driven test automation

• Apply LLMs across the full test lifecycle: test generation from specs and PRs, self-healing UI tests, synthesis, visual regression with vision models, and AI-assisted failure triage

• Validate every AI-generated test through evals — no LLM-authored test ships without proof it doesn't degrade signal

• Establish the eval discipline for Lila's AI/agent stack: golden datasets, rubrics, regression suites, offline + online evaluation pipelines

Define and operate the quality metrics system

• Define quality SLOs and adoption metrics by team and service: coverage, escape rate, MTTR, change-failure rate, eval pass rate, lab QC violation rate

• Build dashboards that make quality visible from PR to executive review

• Apply Google SRE practices to prioritize where investment goes

Mid-long term - Stand up the QC framework for lab automation

• Design the validation framework, harnesses, and contracts that lab and Science Ops teams will operate

• Embed ALCOA+ principles: data integrity, audit trails, lineage from sample → instrument → output

• Partner with Research Ops on pre-flight, in-flight, and post-flight validation patterns for autonomous lab execution

Lead and coach across the engineering org

• Build a 3–5 person team of test automation engineers focused on platform leverage, not on writing tests for other teams

• Coach engineering teams on test design, quality investments, and adoption — make it cheaper to test well than to ship blind

• Translate UX and customer issues into testable contracts and platform improvements

First 6–12 Month Outcomes

• First 90 days: Establish baselines — flake rate, time-to-signal, change-failure rate, coverage, and current build-vs-buy footprint — and publish a quality scorecard with the first set of SLOs. Hire or onboard the initial 1–2 platform engineers.

• By 6 months: Ship v1 of the test automation platform adopted by at least one flagship engineering team by default; land CI/CD test-execution improvements (parallelization, affected-only selection, flake quarantine) with measurable time-to-signal reduction. Stand up the eval discipline…