ML Platform (Cortex)

• Tech Lead for ML Feature Management, a team of 10 engineers.

• Built two generations of the Twitter Feature Store product. Collaborated with teams across the company to facilitate migration and adoption.

• Filled in as senior TL for the ML Data group, 20 engineers across two teams.

Scala, Java, Python, SQL, Jupiter Notebook, Scalding/Hadoop, Airflow, BigQuery, Beam/Dataflow, Thrift, Avro, GCP

Personalization Data and Controls

• Built the data pipelines and application service for surfacing all data Twitter collected for a user, logged in or logged out.

• Worked with Legal to implemented DAA & GDPR personalization controls (user's ability to opt out of personalized ads).

Identity Bridging

• Tech lead for a team of 3.

• Data pipelines for bridging user identifiers for advertising and personalization.

Scala, Java, Scalding/Hadoop, Aurora/Mesos, Thrift

Infrastructure

• AWS-based infrastructure for real-time bidding, user action tracking, data collection.

Data Platform and Reporting

• Job launching framework, Hadoop infrastructure and data pipelines for big scale machine learning, identity graph construction, and report generation.

Python, Java, Cascading/Hadoop, Hive, MySQL, Redis, Thrift, AWS

Application Performance Management

• Designed and developed software for heap dump analysis and memory leak detection through graph mining.

• Integrated acquired products into Enterprise Manager Grid Control.

Application Performance Analytics

• Baseline and Correlation tools SDK.

• Analytics web app: ETL and front end.

Java, PL/SQL, J2EE, Oracle ADF