WhyLabs, Co-Founder and Engineering Lead
Andy Dang is the co-founder and head of engineering at WhyLabs, the AI Observability company on a mission to build the interface between AI and human operators.. Prior to WhyLabs, Andy spent half a decade at Amazon, where he built massive data pipelines for the Advertising Platform, deployed some of the company’s first ML applications to production, built the internal Machine Learning Platform and helped launch the first iterations of SageMaker. Andy holds a Masters in CS from Tokyo Institute of Technology. He is a frequent speaker on topics ranging from MLOps to building responsible AI systems.
Beyond Prod: Don't Let Your Data Fail You
In the era of microservices, decentralized ML architectures and complex data pipelines, data quality has become a bigger challenge than ever. While infrastructure-as-code and DevOps frameworks such as Pulumi enable best practices in managing and testing the infrastructure and software, much is left to be desired for managing data quality. As data becomes more entangled in software-based decisions, it’s critical for companies to start treating data with similar rigor to what the DevOps world has. In this talk, we will address this challenge through whylogs, an open source standard for data logging. We’ll deep-dive how whylogs fit into the general infrastructure as a whole and how it can enable end-to-end observability and monitoring for your data stack. This shift in paradigm will enable companies that operate with data to move faster and safer by building discipline and processes around data.
Scalable monitoring for real time data & ML systems
“Garbage in, garbage out” - is still true for data applications in 2021. In fact, this age-old saying still poses one of the biggest challenges to the data and ML applications. Debugging, troubleshooting & monitoring for data-related bugs takes over the majority of an engineer's day. In DevOps, software operations are taken to a level of an art. Sophisticated tools enable engineers to quickly identify and resolve issues, continuously improving software stability and robustness. In the data world, operations are still largely a manual process that involves Jupyter notebooks and SQL scripts. One of the cornerstones of the DevOps toolchain is logging. Traces and metrics are built on top of logs enabling monitoring and feedback loops. What does logging look like in a real time data and ML system?
In this talk we will show you how to enable statistical data logging for a data application. We will discuss how something so simple enables testing, monitoring and debugging of the entire data pipeline. We will dive deeper into key properties of a logging library that can handle TBs of data in a real time system with Apache Kafka, and how we can enable monitoring at scale of the modern data stack.