First Sessions and Speakers Announced for ODSC West 2023

ODSC West may be several months away still, but we’ve already lined up exciting talks, workshops, and training sessions from some of the leading experts in AI and data science. Check them out below. 

Driving Success for Sellers by Infusing AI in CRM Platform

Sarah Kefayati｜Associate Principal Data Scientist｜IBM

This session will explore how AI algorithms can generate better recommendations by analyzing customer preferences, behavior and purchase history. You’ll also review the architecture and design components that enable personalized recommendations to be integrated into our CRM platform.

MLOps: Monitoring and Managing Drift

Oliver Zeigermann｜Machine Learning Architect｜Freelancer

This workshop will discuss the metrics that can be used as a surrogate to understand the performance of your model. In Part 1 of this session, you will: 

• Simulate production on an existing machine learning model and detecting drift
• Utilize OpenAPI machine learning service 
• We will use Evidently, Prometheus and Grafana to monitor and detect the drift

Part II will cover: 

• Interpreting and Analyzing Drift and what to do about it
• Interpreting what happened to cause the drift and deciding what to do about it 
• You can take is to retrain your model with new data
• Considering rethinking the model architecture or the data we are using

Capturing CAP in a Kappa Data Architecture

Joep Kokkeler｜Senior Data Engineer｜Dataworkz NL

The CAP theorem states that you can not have consistency, availability, and partition tolerance at the same time, but what if choosing for Kappa architecture makes it possible to have it all? This session will discuss what Kappa is and how it compares to Lambda, a microservice and monolithic architecture. 

A Semi-Supervised Anomaly Detection System Through Ensemble Stacking Algorithm

Chuying Ma｜Senior Data Scientist｜Walmart

This session will explore a systematic, flexible, extensible, and holistic anomaly detection architecture to augment the existing labels and detect anomalies with a low cost. The new system is able to incorporate both traditional machine learning models and deep learning-based anomaly detection models to generate a unified anomaly score by the ensemble stacking algorithm to address different types of anomalies simultaneously. 

Personalizing LLMs with a Feature Store

Jim Dowling｜CEO｜Hopsworks

This session will teach you how to personalize LLMs using a feature store and prompt engineering. You’ll see how the open-source feature store, Hopsworks, can be used to build a personalized LLM application. Specifically, you will look at:

• How to build templates for prompts, and how they can be easily constructed and included in user queries
• How to fill in prompt templates with real-time context data, produced by streaming feature pipelines, and user-specific data, produced by batch feature pipelines
• How we can incorporate documents from vector databases in prompts using a combination of user-input and historical user data from the feature store

What is a Time-Series Database and Why Do I Need One?

Jeff Tao｜Founder & CEO｜TDengine

The increase in the size of time-series data sets poses a major challenge for general database management systems like relational and NoSQL databases. This session will cover purpose-built time-series databases, a solution to this challenge. They are much more efficient in terms of ingestion rate, query latency, and data compression. They also include special analytic functions and data management features so that you can develop applications more easily.

Evaluation Techniques for Large Language Models

Rajiv Shah, PhD｜Machine Learning Engineer｜Hugging Face

In this session, you’ll learn about the existing research on the capabilities of LLMs versus small traditional ML models. You’ll also discuss If an LLM is the best solution, and several techniques, including evaluation suites like the EleutherAI Harness, head-to-head competition approaches, and using LLMs for evaluating other LLMs. The tutorial will also touch on subtle factors that affect evaluation, including role of prompts, tokenization, and requirements for factual accuracy. Finally, a discussion of model bias and ethics will be integrated into the working examples.

Data Science Applied to Manufacturing Problems

Angad Arora｜Manufacturing Data Scientist｜Google

This session will discuss how data science can help achieve 3 manufacturing KPIs: Produce more, be efficient and optimize resource utilization, and ship with the highest possible quality. You’ll take a deep dive into examples of key projects that have influenced these KPIs.

What’s next?

Join us at ODSC West 2023 this November to learn more about these topics, and much, much more. Plus, you’ll save 70% on your in-person or virtual pass as part of our Super Early Bird Sale when you register by Friday.