Everest Lab Annual Meeting 2025

Date

Friday, November 14, 2025

Time

8:30 AM – 6:00 PM EST

Location

MIT Building 45, 8th Floor
51 Vassar Street, Cambridge, MA

Format

In-Person Event

Agenda

8:30-9:00 AM

Coffee and Light Breakfast

9:00-9:10 AM

Intros and Welcome

Mike Cafarella, Tim Kraska, Sam Madden

Session 1 – AI Infrastructure 1

9:10–9:25 AM

Continual Learning in practice [and in theory]

Murali Narayanaswamy, Amazon

9:25–9:40 AM

State of Palimpzest Project & Demo

PZ Team, Matthew Russo, Gerardo Vitagliano, Mike Cafarella, Sam Madden, Tim Kraska

9:40–9:55 AM

Log Augmented Generation

Peter Baile Chen

9:55–10:10 AM

State of DSPy Project

Omar Khattab

Session 2 – AI Programming and D4

10:10–10:25 AM

Project Overview and Demo

D4 Team, Jason Mohoney, James Moore, Amadou Ngom, Ziniu Wu, and Alex Zhang, Tim Kraska and Sam Madden

10:25–10:50 AM

Uplift/Downlift

James Moore

10:50–11:05 AM

Self-correcting code generation

Amadou Ngom

11:05–11:20 AM

Break

Session 3 – ML For Systems

11:20–11:35 AM

Scaling to Meet the Needs of Agentic AI

Pradeep Dubey, Intel

11:35–11:50 AM

Automated Data Infrastructure Management with Brad

Geoffrey Yu

11:50–12:05 PM

Robust Query Optimization

Ziniu Wu

12:05-1:05 PM

Lunch

Session 4 – AI Infrastructure 2

1:05-1:20 PM

Agentic Debugging

Sam Madden

1:20-1:35 PM

Next Generation Vector Databases

Sylvia Zhang

Session 5 - AI for Data Science

1:35 - 1:50 PM

Disrupting Finance with AI

HP Bunaes, TWG

1:50-2:05 PM

KramaBench

Gerardo Vitagliano

2:05-2:20 PM

Optimizable AI Systems

Peter Baile Chen

2:20-2:35 PM

Carnot: Deep Research w/Cost-Based Optimization

Matthew Russo

2:35-2:50 PM

Causal Compression

Anna Zeng

2:50-3:00 PM

Break

Session 6 - Multimodal AI

3:00-3:15 PM

Anthony Karthik, Accenture

3:15-3:30 PM

Learned Video Search

Darryl Ho

3:30-3:45 PM

Materials Science Discovery

Eugenie Lai

3:45-4:00 PM

AI for Manufacturing and Design

Mike Cafarella

4:00-4:45 PM

Brainstorming and next steps

4:45-6:00 PM

Reception and Poster Session

Posters

Deep Research is the New Analytics System: Towards Building the Runtime for AI Driven Analytics

Matthew Russo

KramaBench: A Benchmark for AI Systems on Data-to-Insight Pipelines over Data Lakes

Gerardo Vitagliano

Don't Repeat Yourself: A Lesson-Driven Agentic System for Error-Free Data Science

Sushrut Borkar

D4X: Design-Doc Driven, Self-Designing Agentic Systems For Long-Horizon, Complex Tasks

Amadou Latyr Ngom

From Logs to Causal Inference: Diagnosing Large Systems

Markos Markakis

DejaVid: Encoder-Agnostic Learned Temporal Matching for Video Classification

Darryl Ho

Blueprinting the Cloud: Unifying and Automatically Optimizing Cloud Data Infrastructures with BRAD

Geoffrey Yu

Auto-Prep: Holistic Prediction of Data Preparation Steps for Business Intelligence

Eugenie Lai

BackTranslation: Learning to Generate Code Repositories

James Moore

Prism: Synchronized Views for Collaborative Software Development

Jason Mohoney

Archi: Knowledge Discovery and Retrieval at CERN

Luca Lavezzo, Pietro Lugato

SmartVSS: Towards A Unified System for Reasoning-intensive Retrieval Tasks

Sylvia Zhang

Beyond Vector Search: Self-Designing Semantic Indexing for Natural Language Queries

Zhuohan Gu, Tianyu Li

High Performance Buffer Management for Vector Search

Xinjing Zhou

Causal DAG Summarization

Anna Zeng