The Lakehouse Decade: Where Storage Meets Intelligence
A decade-long look at how open table formats reshaped the modern data stack and what comes after the lakehouse.
Dr. Amara Okafor San Francisco
Jul 22, 2025 · 12:30 PMThe program
Every talk, one place
A curated archive of sessions delivered across our US editions — filter by track to find your next deep dive.
Serving Trillion-Token Models on a Budget
Practical patterns for batching, quantization, and autoscaling LLM inference without burning your cloud bill.
Rohan Mehta Austin
Jul 23, 2025 · 3:00 PMExactly-Once at a Million Events per Second
Lessons from running mission-critical streaming pipelines with strict ordering and replay guarantees.
Lin Zhao Chicago
Jul 17, 2024 · 3:00 PMFinOps for Multi-Cloud: Stop Guessing, Start Governing
A governance framework that gives finance and engineering one shared, real-time view of spend.
Erik Larsen Seattle
Jul 16, 2024 · 12:30 PMFoundation Models as Infrastructure
Why the next platform layer is model-shaped, and how to build a reusable ML platform around it.
Dr. Valentina Cruz New York
Jul 18, 2023 · 9:00 AMVectorized Query Engines from First Principles
A ground-up tour of columnar execution, SIMD, and the design tradeoffs behind modern OLAP engines.
Marcus Bell Boston
Jul 17, 2023 · 12:30 PMData Contracts in Practice
How explicit producer–consumer contracts cut pipeline incidents by 60% across hundreds of teams.
Dr. Amara Okafor Denver
Jul 11, 2022 · 2:30 PMReal-Time Feature Stores Without the Pain
Bridging the online/offline gap with a single source of truth for features and labels.
Lin Zhao Atlanta
Jul 12, 2022 · 10:30 AMThe Warehouse Is the New Operating System
From dashboards to applications: how the cloud warehouse became a programmable platform.
Marcus Bell Las Vegas
Jul 20, 2021 · 10:30 AMBuilding Feature Platforms That Scale to Thousands of Models
A blueprint for feature stores that stay fast, consistent, and debuggable as model count explodes.
Priya Nair San Francisco
Jul 22, 2025 · 9:00 AMConsensus at Planet Scale
What it really takes to run replicated, strongly-consistent storage across continents.
Kenji Tanaka San Francisco
Jul 21, 2025 · 11:00 AMCausal Inference for Product Decisions
Moving teams beyond correlation dashboards to causal models that actually inform decisions.
Sofia Russo Seattle
Jul 15, 2024 · 10:30 AMQuery Optimization in the Age of Open Formats
How cost-based optimizers are being rebuilt for Iceberg, Delta, and the disaggregated stack.
David Goldberg Seattle
Jul 15, 2024 · 2:30 PMEvaluating LLMs You Can Actually Trust
A practical framework for building evaluation suites that catch real regressions, not vibes.
Dr. Aisha Rahman Seattle
Jul 17, 2024 · 9:00 AMGolden Paths: Platform Engineering on Kubernetes
Designing internal developer platforms that abstract Kubernetes without hiding what matters.
Carlos Mendez New York
Jul 19, 2023 · 9:00 AMReal-Time Recommenders at Consumer Scale
Serving large embedding models and fresh features with single-digit-millisecond latency.
Mei Chen New York
Jul 18, 2023 · 2:30 PMObservability for Data Pipelines
Lineage, freshness, and anomaly detection — the telemetry that keeps data products reliable.
James O'Brien Atlanta
Jul 13, 2022 · 10:30 AMFederated Learning Without the Hype
Where privacy-preserving training genuinely pays off — and where a central warehouse still wins.
Dr. Fatima Al-Sayed Atlanta
Jul 11, 2022 · 4:30 PMData Quality as a First-Class Product
Treating freshness, completeness, and contracts as SLAs your consumers can depend on.
Wei Liu Las Vegas
Jul 19, 2021 · 1:00 PMThe Road to the Lakehouse
A look back at how warehouses and lakes converged, and why open formats won the decade.
Marcus Bell Virtual
Jul 15, 2020 · 1:00 PMScaling ML Platforms When Everyone Went Remote
Lessons from operating shared ML infrastructure for distributed teams overnight.
Priya Nair Virtual
Jul 16, 2020 · 12:30 PMSchema Evolution in Open Table Formats
Safe, backward-compatible schema change strategies for petabyte-scale tables.
Dr. Amara Okafor Chicago
Jul 16, 2019 · 12:30 PMElastic Compute for Bursty Analytics
Autoscaling analytical workloads without paying for idle capacity between the bursts.
Erik Larsen Las Vegas
Jul 12, 2018 · 11:00 AMFoundations of the Cloud Data Warehouse
The opening keynote that launched the conference: separating storage from compute, for good.
Dr. Valentina Cruz San Jose
Sep 19, 2017 · 1:00 PMReal-Time Sync: Event-Driven Integration Across Enterprise Systems
Replacing brittle scheduled ETL with change-data-capture and streaming so account updates propagate to every downstream system in seconds, not hours.
Srikanth Jonnakuti San Francisco
Jul 23, 2025 · 9:00 AMSelf-Serve Data Platforms That Don't Break Trust
How governed, self-serve platforms let analysts ship fast while keeping lineage and access under control.
Naomi Clarke San Francisco
Jul 22, 2025 · 11:00 AMTail Latency Is the Product: Serving LLMs in Real Time
Batching, speculative decoding, and scheduling tricks that keep p99 latency low under bursty load.
Arjun Patel San Francisco
Jul 22, 2025 · 11:00 AMExperimentation at Scale Without Fooling Yourself
Designing A/B systems that account for interference, peeking, and the messy reality of product metrics.
Hana Kim San Francisco
Jul 21, 2025 · 1:00 PMEmbeddings in Production: Retrieval That Stays Fresh
Keeping vector indexes consistent and low-latency as embeddings and documents change constantly.
Lucia Fernandez San Francisco
Jul 21, 2025 · 2:30 PMReading the Cloud Bill: FinOps for Engineers
Turning raw billing exports into engineering decisions developers can actually act on.
Ben Fisher San Francisco
Jul 21, 2025 · 2:30 PMGolden Paths for Production ML
Making the common ML workflows boring: paved roads that turn heroics into routine releases.
Anita Desai San Francisco
Jul 21, 2025 · 9:00 AMChange Data Capture at High Volume
Keeping operational databases and the lakehouse in sync without melting the source system.
Omar Haddad Seattle
Jul 16, 2024 · 10:30 AMOn-Call for Data: Building a Reliability Culture
The tooling, runbooks, and incentives that keep data products dependable and engineers rested.
Terrence Walker Seattle
Jul 16, 2024 · 3:00 PMBenchmarks You Can Trust
Why most ML benchmarks mislead, and how to build evaluation suites that survive contact with reality.
Dr. Katherine Wolfe Seattle
Jul 17, 2024 · 4:30 PMStorage Engines for Geo-Distributed Databases
Design tradeoffs in LSM trees, replication, and consensus for databases that span continents.
Daniel Park Seattle
Jul 17, 2024 · 9:00 AMSecure Multi-Cloud Networking for Data Teams
Connecting data platforms across providers without opening holes your security team will hate.
Diego Ramirez Seattle
Jul 17, 2024 · 10:30 AMFairness Guardrails That Ship
Translating fairness research into measurable guardrails teams can actually put into production.
Dr. Grace Adeyemi New York
Jul 17, 2023 · 3:00 PMThe Modern Ingestion Stack
Batch, streaming, and CDC under one roof — choosing the right tool for each source without sprawl.
Naomi Clarke New York
Jul 19, 2023 · 1:00 PMCausal Measurement for Growth Teams
Geo experiments, switchbacks, and synthetic controls when a clean A/B test just isn't possible.
Hana Kim New York
Jul 18, 2023 · 12:30 PMScaling the ML Platform Team
Org design, interfaces, and the platform-as-product mindset behind ML infrastructure that scales.
Anita Desai New York
Jul 17, 2023 · 4:30 PMPipeline Reliability SLAs in Practice
Defining freshness and completeness SLAs your consumers can plan around — and meeting them.
Terrence Walker Atlanta
Jul 13, 2022 · 2:30 PMStreaming Ingestion Without the 3 a.m. Pages
Backpressure, schema drift, and dead-letter handling for ingestion pipelines that stay calm.
Omar Haddad Atlanta
Jul 11, 2022 · 1:00 PMVector Search Before It Was Cool
Hard-won lessons building approximate nearest-neighbor systems for recommendations at scale.
Lucia Fernandez Atlanta
Jul 11, 2022 · 10:30 AMCost Optimization for the Lakehouse
Right-sizing compute, tuning file layouts, and killing the silent spend in your data platform.
Ben Fisher Atlanta
Jul 13, 2022 · 10:30 AMEvaluating Models When Everyone Works Remote
Reproducible, shareable evaluation pipelines for teams that no longer sit in the same room.
Dr. Katherine Wolfe Virtual
Jul 20, 2021 · 2:30 PMConsensus Protocols, Explained for Builders
A practical tour of Raft, Paxos variants, and the failure modes that bite real deployments.
Daniel Park Virtual
Jul 19, 2021 · 2:30 PMMulti-Cloud by Necessity, Not Hype
When spanning providers genuinely pays off — and the networking realities nobody warns you about.
Diego Ramirez Virtual
Jul 15, 2020 · 1:00 PMBuilding Governed Platforms Overnight
Lessons from standing up governed, remote-first data platforms when the office disappeared.
Naomi Clarke Virtual
Jul 16, 2020 · 3:00 PMInference Economics in the Early GPU Crunch
Batching and model selection strategies from the days when every GPU hour really hurt.
Arjun Patel Chicago
Jul 18, 2019 · 12:30 PMReliability Lessons from a Decade of Outages
What recurring incidents taught one team about lineage, alerting, and blameless postmortems.
Terrence Walker Chicago
Jul 18, 2019 · 10:30 AMElastic Storage for the Bursty Enterprise
Decoupling storage growth from compute cost when workloads are spiky and unpredictable.
Ben Fisher Las Vegas
Jul 10, 2018 · 1:00 PMEarly Experimentation Platforms
Building the first in-house A/B platform: metrics, guardrails, and the trust it took to earn.
Hana Kim Las Vegas
Jul 11, 2018 · 10:30 AMMeasuring Models in the Warehouse Era
An early call for rigorous, reproducible model evaluation as data moved into the cloud warehouse.
Dr. Grace Adeyemi San Jose
Sep 19, 2017 · 11:00 AMDesigning Storage for the New Warehouse
The storage-engine decisions behind separating compute from durable, columnar data.
Daniel Park San Jose
Sep 21, 2017 · 11:00 AMIntelligent Automation for Cloud-Native Integration Platforms
How model-assisted automation reduced manual reconciliation and incident toil across a multi-account, event-driven integration estate.
Srikanth Jonnakuti San Francisco
Jul 22, 2025 · 3:00 PMSecure DevSecOps for Multi-Account Cloud Infrastructure
Policy-as-code, automated guardrails, and infrastructure templates that keep enterprise cloud estates compliant without slowing delivery.
Diego Ramirez San Francisco
Jul 21, 2025 · 4:30 PM