Joined the newly formed Meta Superintelligence Labs to build foundational systems for AI Trust & Safety.

Focused on creating and scaling the core safety platform, enabling product teams to integrate robust safety and integrity guardrails across all of Meta's current and future AI products.

• Led the architectural design and unification of proactive and reactive Messenger Wellbeing

ML platforms for harm detection.

• Engineering lead for unified platform to manage feature dependencies and enable feature unification across proactive and reactive systems.

• Drove system operations maturity by establishing and implementing SLIs/SLOs across the entire Messenger Wellbeing organization.

• Enhanced Messenger Search integrity features to better protect users from harmful content and interactions.

Over a four-year tenure in the Messenger Wellbeing organization, I led the re-architecture of the integrity review platform, delivering significant improvements in scalability, self-service onboarding for new use cases, and system immutability.

Enhanced human review effectiveness by closing critical enforcement gaps and integrating new signals to better tackle high-severity harms like CSE and CSAM. I implemented and scaled Meta's central ML platform for Messenger, building numerous ML features to power harm detection models.

This work had a massive, direct impact on user safety, scaling our enforcement capacity to millions of reviews (reactive & proactive) daily. This resulted in millions of harmful interactions being caught & enforced, or prevented.

Senior engineer responsible for code re-engineering, performance profiling and optimisation for an Apache Spark-based VAR calculation engine in early stages of development.

Delivered significant code simplification through appropriate abstraction layers, utilising open-source frameworks and reducing code duplication.

Realised significant performance gains by introducing a systematic approach to performance analysis, including performance metric capture, performance profiling, analysis of optimisation potential, and implementation of performanceimprovements. Reduced the batch execution from multiple days to several hours.

Lead development on initial cloud migration of the platform utilising Gitlab, Terraform and Dataproc on Google Cloud.

Now leading the full migration of the platform and reviewing opportunities to migrate to cloud-native technologies including BigQuery, BigTable, MemoryStore, Cloud Functions and Cloud Endpoints.

Key Technologies: Scala, Java, Apache Spark, Hive, Terraform, Docker, Kubernetes, Google Cloud - GCP