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

Flink applications processing millions of messages per second, built in days with 3x smaller client code and 4x lower time to prod than the Datastream API and a fully testable, CI/CD enabled lifecycle to boot - what’s not to like! Come listen to how Pinterest systematically enables developing Flink applications at scale. We will talk (for the first time) about our in-house framework - Near Realtime Galaxy (NRTG) - that builds composable, large-scale, dependency injection based stream applications and the platform it enables. We will discuss our experiences while building NRTG on Flink - happy topics like Serialization, Async RPCs, Error Recovery, Late Data, Backfills and demo our dev to prod process: a CI/CD based lifecycle with automated checks, health monitoring and minimal manual intervention. 0:00 Pinterest Architecture & Business Requirements 3:40 What is NRTG, Why NRTG 9:30 The Fast, Systematics and Scalable Promise 21:13 Onboarding Outcomes 22:56 CI/CD in the Future

Demystifying Deployments: Applications or Clusters, Active and Reactive Scaling - What is it about?
▶︎

Demystifying Deployments: Applications or Clusters, Active and Reactive Scaling - What is it about?

Massive Scale Data Processing at Netflix using Flink - Snehal Nagmote & Pallavi Phadnis
▶︎

Massive Scale Data Processing at Netflix using Flink - Snehal Nagmote & Pallavi Phadnis

Enabling Apache Flink Management Platforms as a Service in the AI Stack
▶︎

Enabling Apache Flink Management Platforms as a Service in the AI Stack

Apache Fluss and the Seven Deadly Sins of Streaming
▶︎

Apache Fluss and the Seven Deadly Sins of Streaming

Iran – Neuseeland Highlights | Gruppe G, FIFA WM 2026 | sportstudio
▶︎

Iran – Neuseeland Highlights | Gruppe G, FIFA WM 2026 | sportstudio

Turing Award Winner: Disagreeing with Google, Postgres, Future Problems | Mike Stonebraker
▶︎

Turing Award Winner: Disagreeing with Google, Postgres, Future Problems | Mike Stonebraker

Apache Fluss: Making Your Lakehouse Truly Real Time
▶︎

Apache Fluss: Making Your Lakehouse Truly Real Time

Stop Rambling: The 3-2-1 Speaking Trick That Makes You Sound Like A CEO
▶︎

Stop Rambling: The 3-2-1 Speaking Trick That Makes You Sound Like A CEO

Preventing Blackouts with Flink: Power Grid Data Processing for Millisecond-Level Fault Handling
▶︎

Preventing Blackouts with Flink: Power Grid Data Processing for Millisecond-Level Fault Handling

Let’s Handle 1 Million Requests per Second, It’s Scarier Than You Think!
▶︎

Let’s Handle 1 Million Requests per Second, It’s Scarier Than You Think!

Andrej Karpathy: From Vibe Coding to Agentic Engineering w/ Stephanie Zhan
▶︎

Andrej Karpathy: From Vibe Coding to Agentic Engineering w/ Stephanie Zhan

Charting New Territory: LinkedIn’s Early Bet on Flink Batch for Large-Scale Workloads
▶︎

Charting New Territory: LinkedIn’s Early Bet on Flink Batch for Large-Scale Workloads

LIVE: Conan O’Brien speaks at Harvard graduation ceremony (full)
▶︎

LIVE: Conan O’Brien speaks at Harvard graduation ceremony (full)

From Idea to $650M Exit: Lessons in Building AI Startups
▶︎

From Idea to $650M Exit: Lessons in Building AI Startups

Modern Architecture 101 for New Engineers & Forgetful Experts - Jerry Nixon - NDC Copenhagen 2025
▶︎

Modern Architecture 101 for New Engineers & Forgetful Experts - Jerry Nixon - NDC Copenhagen 2025

KEYNOTE: Apache Flink in the Cloud-native Era
▶︎

KEYNOTE: Apache Flink in the Cloud-native Era

How I’m Preparing For The “Supercycle”
▶︎

How I’m Preparing For The “Supercycle”

RAG Crash Course for Beginners
▶︎

RAG Crash Course for Beginners

Flink Forward Barcelona 2025: Opening Session and Keynotes
▶︎

Flink Forward Barcelona 2025: Opening Session and Keynotes

Flink State Management: A Journey from Core Primitives to Next-Generation Incremental Computation
▶︎

Flink State Management: A Journey from Core Primitives to Next-Generation Incremental Computation