Hardened production LLM safety across sensitive policy domains at one of the world's largest AI deployments.

▸ Designed red-teaming and adversarial evaluation frameworks that surfaced model failure modes before deployment — directly improving trust, accuracy, and policy compliance.

▸ Built structured evaluation rubrics aligned to Meta's policy taxonomy; feedback was incorporated into model improvement cycles.

▸ Iterated on prompt-based test suites to reduce false negatives in high-risk content classification, cross-validating against ground-truth benchmarks.

Led a 10-engineer team building the microservices-based localization infrastructure that powered LinkedIn's 800M+ user experience across 15+ global markets.

▸ Sustained 99.8% uptime on all platform services while scaling to support continuous global product launches.

▸ Cut QA cycle time 40% and defect escape rate 35% by re-architecting the translation QA pipeline with automated validation layers and ML-assisted quality scoring.

▸ Delivered 15+ market expansions on schedule by redesigning content ingestion APIs and distributed processing pipelines — reducing per-locale launch overhead ~60%.

▸ Managed $3M annual budget with zero overruns across 6 years; reallocated 20% of vendor spend to internal tooling that eliminated two external dependencies.

▸ Grew 3 ICs to Staff/Senior through structured roadmap ownership and cross-functional sponsorship.

Designed and built the automation infrastructure and APIs that became the foundation for LinkedIn's global localization platform.

▸ Saved thousands of engineering hours annually by building end-to-end automation pipelines that eliminated manual translation handoffs across 15+ markets.

▸ Reduced i18n defect rate 40% by authoring static analysis tools that caught locale-specific encoding and formatting bugs at CI time.

▸ Designed the REST/event-driven interfaces that downstream teams still rely on to submit, track, and retrieve localized content at scale.

• Conducted manual and automated testing on software applications to ensure functionality and reliability.

• Analyzed defects to determine corrective action plans and created detailed quality assurance reports.

• Reviewed test cases and tracked bugs using bug tracking systems like JIRA.

• Submitted localization bugs and worked on engineering tasks using JIRA

• Developed Selenium based functional test automation scripts in Ruby

• Troubleshot automation scripts and participated in Scrum team meetings