MLOps Community
timezone
+00:00 GMT
Livestream
LLMs in Production - Part II
# LLM in Production
# Large Language Models

Large Language Models in Production Conference

Join us for two days of talking with some of our favorite people at the forefront of using LLMs in the wild, and an in-person workshop in San Francisco on how to build and deploy LLM based apps hosted by Anyscale.

There will be over 50 Speakers from Stripe, Meta, Canva, Databricks, Anthropic, Cohere, Redis, Langchain, Chroma, Humanloop and so many more.

This all started after we put together the LLM in-production survey and realized there are still lots of unknowns when dealing with LLMs, especially when dealing with them at scale. We open-sourced all the responses and we decided if no one was going to talk about working with LLMs in a non-over-hyped way, we would have to.

Let's discover how to use these damn probabilistic models in the best ways possible without sacrificing the necessary software design building blocks.

Expect all the fun and learnings from the first one. DOUBLED.

And remember, there will be some sweeeet sweet swag giveaways.

Huge Shoutout to all our sponsors of this event, find more info about them below.

Speakers
Matei Zaharia
Matei Zaharia
Cofounder and Chief Technologist @ Databricks
Chip Huyen
Chip Huyen
CEO @ Claypot AI
Sarah Aerni
Sarah Aerni
Vice President, AI/Machine Learning and Engineering @ Salesforce
Emmanuel Ameisen
Emmanuel Ameisen
Research Engineer @ Anthropic
Sumit Kumar
Sumit Kumar
Senior Machine Learning Engineer @ Meta
Shreya Rajpal
Shreya Rajpal
Creator @ Guardrails AI
Aravind Srinivas
Aravind Srinivas
CEO & Co-Founder @ Perplexity.ai
Ines Chami
Ines Chami
Co-Founder @ NumbersStationAI
Samyam Rajbhandari
Samyam Rajbhandari
Principal Architect @ Microsoft Corporation
Omar Sanseviero
Omar Sanseviero
Machine Learning Lead @ Hugging Face
Alex Ratner
Alex Ratner
CEO and Co-founder @ Snorkel AI
Sam Charrington
Sam Charrington
Host @ TWIML AI Podcast
Raza Habib
Raza Habib
CEO and Co-founder @ Humanloop
Chris Van Pelt
Chris Van Pelt
Co-founder / CISO @ Weights & Biases
Samuel Partee
Samuel Partee
Principal Applied AI Engineer @ Redis
Sophie Daly
Sophie Daly
Staff Data Scientist @ Stripe
Misty Free
Misty Free
Product Manager @ Jasper
Scott Mackie
Scott Mackie
Founding Engineer @ Mem Labs
Anton Troynikov
Anton Troynikov
Founder / Head of Technology @ Chroma
Sahar Mor
Sahar Mor
Product Lead @ Stripe
George Mathew
George Mathew
Managing Director @ Insight Partners
Manjot Pahwa
Manjot Pahwa
VP @ Lightspeed
Azin Asgarian
Azin Asgarian
AI Technical Lead @ Georgian
Lance Martin
Lance Martin
Software engineer @ LangChain
Daniel Campos
Daniel Campos
Research Scientist @ Snowflake
Richa Sachdev
Richa Sachdev
Executive Director- Data Operations and Automation @ JP Morgan Chase
Chris Brousseau
Chris Brousseau
Lead Data Scientist - International Focus @ Mastercard
Xin Liang
Xin Liang
Senior Machine Learning Engineer @ Canva
Denys Linkov
Denys Linkov
ML Lead @ Voiceflow
Monmayuri Ray
Monmayuri Ray
MLOPS Advisor @ Gitlab
Joseph Gonzalez
Joseph Gonzalez
Professor, Co-Founder & VP of Product @ UC Berkeley, Aqueduct
Amrutha Gujjar
Amrutha Gujjar
CEO & Co-Founder @ Structured
Daniel Vila Suero
Daniel Vila Suero
CEO & Co-Founder @ Argilla
David Aponte
David Aponte
Senior Research SDE, Applied Sciences Group @ Microsoft
Travis Fischer
Travis Fischer
Founder / CEO @ Stealth
Tristan Zajonc
Tristan Zajonc
Cofounder & CEO @ Continual
Vikram Sreekanti
Vikram Sreekanti
CEO @ Aqueduct
Daniel Whitenack
Daniel Whitenack
Founder and Data Scientist @ Prediction Guard
Manojkumar Parmar
Manojkumar Parmar
CEO, CTO @ AIShield - Corporate Startup of Bosch
Willem Pienaar
Willem Pienaar
Co-Founder & CTO @ Cleric
Abi Aryan
Abi Aryan
Machine Learning Engineer @ Independent Consultant
Soham Chatterjee
Soham Chatterjee
Machine Learning Lead @ Sleek
Josh Tobin
Josh Tobin
Founder @ Gantry
Sohini Roy
Sohini Roy
Senior Developer Relations Manager @ NVIDIA
Sebastian Cattes
Sebastian Cattes
Senior Data Scientist @ INWT Statistics
Aparna Dhinakaran
Aparna Dhinakaran
Co-Founder and Chief Product Officer @ Arize AI
Gerred Dillon
Gerred Dillon
Unicorn Engineer @ Defense Unicorns
Andrew Mauboussin
Andrew Mauboussin
Engineering Lead @ Surge AI
Asmitha Rathis
Asmitha Rathis
Machine Learning Engineer @ PromptOps
Rohit Agarwal
Rohit Agarwal
CEO @ Portkey.ai
Oscar Rovira
Oscar Rovira
Co-founder @ Mystic AI
Mark Huang
Mark Huang
Co-Founder @ Gradient
Mark Craddock
Mark Craddock
Founder & CTO @ FirstLiot Ltd.
Shrinand Javadekar
Shrinand Javadekar
All things Kubernetes @ Outerbounds, Inc.
Alberto Rizzoli
Alberto Rizzoli
CEO @ V7
Maxime Beauchemin
Maxime Beauchemin
CEO and Founder @ Preset
Chunting Zhou
Chunting Zhou
Research Scientist @ Meta AI
Claire Longo
Claire Longo
Head of ML Solutions Engineering @ Arize AI
Emmy Li
Emmy Li
Technical Trainer @ Anyscale
Raahul Dutta
Raahul Dutta
MLOps Lead @ Elsevier
Bradley Heilbrun
Bradley Heilbrun
Engineer @ Replit
Demetrios Brinkmann
Demetrios Brinkmann
Chief Happiness Engineer @ MLOps Community
Filippo Pedrazzini
Filippo Pedrazzini
CTO @ Prem Labs
Natalia Burina
Natalia Burina
AI Product Leader @ ex Meta AI
Waleed Kadous
Waleed Kadous
Head of Engineering @ Anyscale
Diego Oppenheimer
Diego Oppenheimer
Co-founder @ Guardrails AI
Yada Pruksachatkun
Yada Pruksachatkun
ML Lead @ Moonhub
Stefan Ojanen
Stefan Ojanen
Director of Product Management @ Genesis Cloud
Travis Addair
Travis Addair
CTO @ Predibase Inc.
Rohit Agarwal
Rohit Agarwal
Co--Founder @ Portkey.ai
Vipul Ved Prakash
Vipul Ved Prakash
Co-founder & CEO @ Together
Yujian Tang
Yujian Tang
Developer Advocate @ Zilliz
David Hershey
David Hershey
Vice President @ Unusual Ventures
Adam Breindel
Adam Breindel
Technical Instructor @ Anyscale
Davis Treybig
Davis Treybig
Partner @ Innovation Endeavors
Dina Yerlan
Dina Yerlan
Head of Product, Generative AI Data @ Adobe, Firefly
Rahul Parundekar
Rahul Parundekar
Founder @ A.I. Hero, Inc.
Patrick Barker
Patrick Barker
CTO @ Kentauros AI
Hemant Jain
Hemant Jain
Senior Software Engineer, ML Inference @ Cohere
Atindriyo Sanyal
Atindriyo Sanyal
Co-founder, CTO @ Galileo
Mathieu Bastian
Mathieu Bastian
Director of Engineering, Data Products @ GetYourGuide
Nikunj Bajaj
Nikunj Bajaj
Cofounder & CEO @ TrueFoundry
Artem Harutyunyan
Artem Harutyunyan
Co-Founder & CTO @ Bardeen AI
Travis Cline
Travis Cline
Engineering Manager, Platform @ Virta
Daniel Jeffries
Daniel Jeffries
Managing Director @ AI Infrastructure Alliance
Will Gaviria Rojas
Will Gaviria Rojas
Co-Founder @ Coactive AI
Agenda
Day 1
Day 2
Track View
Track 1
Track 2
Workshops
In-Person Events
3:00 PM
3:30 PM
Stage 1
Opening / Closing
calendar
Welcome & Opening

Plus a little LLM in production survey report tl;dr summary

+ Read More
Demetrios Brinkmann
3:30 PM
4:00 PM
Stage 1
Keynote
calendar
LLMOps: The Emerging Toolkit for Reliable, High-quality LLM Applications

Large language models are fluent text generators, but they often make errors, which makes them difficult to deploy in high-stakes applications. Using them in more complicated pipelines, such as retrieval pipelines or agents, exacerbates the problem. In this talk, Matei will cover emerging techniques in the field of “LLMOps” — how to build, tune and maintain LLM-based applications with high quality. The simplest tools are ones to test and visualize LLM results, some of which are now being incorporated into MLOps frameworks like MLflow. However, there are also rich techniques emerging to “program” LLM pipelines and control LLMs’ outputs to achieve desired goals.

Matei will discuss Demonstrate-Search-Predict (DSP) from my group as an example programming framework that can automatically improve an LLM-based application based on feedback, and other open-source tools for controlling outputs and generating better training and evaluation data for LLMs. This talk is based on their experience deploying LLMs in many applications at Databricks, including the QA bot on our public website, internal QA bots, code assistants, and others, all of which are making their way into our MLOps products and MLflow.

+ Read More
Matei Zaharia
4:00 PM
4:30 PM
Stage 1
Keynote
calendar
Open Challenges For LLM Applications

What do we need to be aware of when building for production? In this talk, we will explore the key challenges that arise when taking an LLM to production.

+ Read More
Chip Huyen
4:30 PM
5:00 PM
Stage 1
Panel Discussion
calendar
Evaluation

Language models are very complex thus introducing several challenges in interpretability. The large amounts of data required to train these black-box language models make it even harder to understand why a language model generates a particular output. In the past, transformer models were typically evaluated using perplexity, BLEU score or human evaluation.

However, LLMs amplify the problem even further due to their generative nature thus making them further susceptible to hallucinations and factual inaccuracies. Thus, evaluation becomes an important concern.

+ Read More
Abi Aryan
Amrutha Gujjar
Josh Tobin
Sohini Roy
5:10 PM
5:20 PM
Stage 1
1:1 networking
calendar
Musical Break

Bring your prompts to the chat cause we will be improvising songs from the audience's suggestions!

+ Read More
Demetrios Brinkmann
5:20 PM
5:50 PM
Stage 1
Presentation
calendar
Using Vector Databases for LLM Part 2: Practical Advice for Production

In the last LLM in Production event, I spoke on some of the ways we've seen people use a vector database for large language models. This included use cases like information/context retrieval, conversational memory for chatbots, and semantic caching.

These are great and make for flashy demos, however, using this in production isn't trivial. Often times, the less flashy side of these use cases can present huge challenges such as: Advice on prompts? How do I chunk up text? What if I need HIPAA compliance? On-premise? What if I change my embeddings model? What index type? How do I do A/B tests? Which cloud platform or model API should I use? Deployment strategies? How can I inject features from my feature platform? Langchain or LlamaIndex or RelevanceAI???

This talk will detail a distillation of a year+ worth of deploying Redis for these use cases for customers and distill it down into 20 minutes.

+ Read More
Samuel Partee
5:50 PM
6:00 PM
Stage 1
Lightning Talk
calendar
Foundation Models in the Modern Data Stack

As Foundation Models (FMs) continue to grow in size, innovations continue to push the boundaries of what these models can do on language and image tasks. This talk will describe our work on applying foundation models to structured data tasks like data linkage, cleaning and querying. We will then discuss challenges and solutions that these models present for production deployment in the modern data stack.

+ Read More
Ines Chami
6:00 PM
6:10 PM
Stage 1
Lightning Talk
calendar
AI Meets Memes: Taking ImgFlip's 'This Meme Does Not Exist' to the next level with a Large Language Model

How to use a Large Language Model (LLM) to create memes? We’ll discuss the unique dataset of ImgFlip, the selection, and fine-tuning of a commercially usable LLM, and associated challenges. Of course, we’ll also demonstrate the model prototype itself. We will also discuss the challenges we anticipate facing with the productionization of an LLM that is used by millions of users.

+ Read More
Stefan Ojanen
6:10 PM
6:20 PM
Stage 1
Lightning Talk
calendar
Building Reliable AI Agents

Autonomous AI agents have gotten a lot of attention recently, but they're mostly just toys. What are the primitives that we need to build more reliable agents, and what are the main business use cases that agentic automation will enable over the next few years?

+ Read More
Travis Fischer
6:20 PM
6:50 PM
Stage 1
Presentation
calendar
Pitfalls and Best Practices — 5 lessons from LLMs in production

Humanloop have now seen hundreds of companies go on the journey from playground to production. In this talk we’ll share case-studies of what has and hasn’t worked. We’ll share what the common pitfalls are, emerging best practices and suggestions for how to plan in such a quickly evolving space.

+ Read More
Raza Habib
6:50 PM
7:10 PM
Stage 1
1:1 networking
calendar
Guided Meditation

Put down the screen for a moment, close your eyes and bliss out in between the sessions

+ Read More
Demetrios Brinkmann
7:10 PM
7:40 PM
Stage 1
Presentation
calendar
Scalable Evaluation and Serving of Open Source LLMs

While we've seen great progress on Open Source LLMs, we haven't seen the same level of progress on systems to serve those LLMs in production contexts. In this presentation, I work through some of the challenges of taking open source models and serving them in production.

+ Read More
Waleed Kadous
7:40 PM
8:10 PM
Stage 1
Panel Discussion
calendar
LLMs on K8s

Large Language Models require a new set of tools... or do they? K8s is a beast and we like it that way. How can we best leverage all the battle hardened tech that k8s has to offer to make sure that our LLMs go brrrrrrr. Lets talk about it in this chat.

+ Read More
Shrinand Javadekar
Manjot Pahwa
Rahul Parundekar
Patrick Barker
8:10 PM
8:40 PM
Stage 1
Presentation
calendar
Embeddings and Retrieval for LLMs: Techniques and Challenges

Retrieval augmented generation with embeddings and LLMs has become an important workflow for AI applications.

While embedding-based retrieval is very powerful for applications like 'chat with my documents', users and developers should be aware of key limitations, and techniques to mitigate them.

+ Read More
Anton Troynikov
8:40 PM
9:00 PM
Stage 1
1:1 networking
calendar
Prompt Injection Game

You think you got prompting skills? been reading too many Reddit threads thinking you can crack the code? Well, let's see what you are capable of!

+ Read More
Demetrios Brinkmann
9:00 PM
9:30 PM
Stage 1
Presentation
calendar
Build and Customize LLMs in Less than 10 Lines of YAML

Generalized models solve general problems. The real value comes from training a large language model (LLM) on your own data and finetuning it to deliver on your specific ML task.Now you can build your own custom LLM, trained on your data and fine-tuned for your generative or predictive task in ten lines of code with Predibase and Ludwig, the low-code deep learning framework developed and open sourced by Uber, now maintained as part of the Linux Foundation. Using Ludwig’s declarative approach to model customization, you can take a pre-trained large language model like LLaMA and tune it to output data specific to your organization, with outputs conforming to an exact schema. This makes building LLMs fast, easy, and economical.In this session, Travis Addair, CTO of Predibase and co-maintainer of open-source Ludwig, will share how LLMs can be tailored to solve specific tasks from classification to content generation, and how you can get started building a custom LLM in just a few lines of code.

+ Read More
Travis Addair
9:30 PM
9:40 PM
Stage 1
Lightning Talk
calendar
Building Production Copilots

Copilots embedded within SaaS applications have become one of the dominant ways of leveraging LLMs within products. In this lightning talk, I’ll review some of the dominant UI paradigms and features, general design patterns and system architectures, and top challenges and future frontiers of production copilot systems.

+ Read More
Tristan Zajonc
9:40 PM
9:50 PM
Stage 1
Lightning Talk
calendar
Building Recommender Systems with Large Language Models

Many researchers have recently proposed different approaches to building recommender systems using LLMs. These methods convert different recommendation tasks into either language understanding or language generation templates. This talk highlights some of the recent work done on this theme.

+ Read More
Sumit Kumar
9:50 PM
9:59 PM
Stage 1
Lightning Talk
calendar
Unleashing Code Completion with LLM's

This would be a talk on the learning on building Code Suggestions, my team has takes in reference to Model, ML Infra, Evaluation to Compute and Cost.

+ Read More
Monmayuri Ray
Sponsors

Diamond

Gold

Silver

Community

Event has finished
June 15, 2:45 PM, GMT
Online
Organized by
MLOps Community
MLOps Community
Event has finished
June 15, 2:45 PM, GMT
Online
Organized by
MLOps Community
MLOps Community