Heng is a staff engineer and tech lead at Pinterest, where he specializes in data ingestion, stream processing, incremental processing and lakehouse architectures at PB scale with open-source technologies like Kafka, Flink, Iceberg and Spark. He has architected scalable systems and critical pipelines aiming to improve user experience and monetization at Pinterest and he is a frequent speaker at various tech conferences on these subjects. He earned his PH.D degree from University of Maryland, College Park with thesis on multimodal machine learning and active authentication on mobile devices.

Recent Talks:
Current 2025 London (May 20)- FlinkSQL Powered Asynchronous Data Processing in Pinterest’s Rule Engine Platform

Current 2025 London (May 21)- Building a scalable and efficient real-time event counter framework with Flink(SQL) and Kafka

Iceberg Summit 2025 - Apache Iceberg as unified storage stack to power batch and stream processing at Pinterest

Current 2023 - Evolution of real-time user engagement event consumption at Pinterest

Flink Forward 2023 - Next generation DB ingestion at Pinterest

Current 2022 - Why Wait? Realtime Ingestion

Flink Meetup May 2022 - Flink-Powered Stream Processing Platform at Pinterest

Flink Forward 2021 - NRTG: Fast, Systematic & Scalable Flink Application Development at Pinterest

Kafka Summit 2021 - Improving Logging Ingestion Quality At Pinterest: Fighting Data Corruption And Tracking Data Loss