MLOps Lightning Talk Night

## Overview

Gather with other machine learning engineering folks in Chicago to hear some 5-10 minute lightning talks about developing and deploying machine learning applications.

This event features:

• Free food (light snacks) & drinks (alcoholic and non-alcoholic)
• Lightning talks about MLOps
• Ability to ask questions to speakers during Q & A
• Opportunities to meet fellow MLOps practitioners

Schedule:

```
5:30pm -> doors open
5:30pm - 6:15pm -> food/drinks/networking
6:15pm - 6:20pm -> introduction from the organizers
6:20pm - 6:50pm -> first 2 talks
6:50pm - 7:00pm -> break
7:00pm - 7:45pm -> last 2-3 talks
7:45pm - 8:00pm -> networking
8:00pm -> doors close
8:10pm -> possible walk to a nearby bar for folks who want to continue mingling
```

## Talks

"Conformal Prediction for Timeseries Forecasting"
speaker: Kevin Kho

In this talk, Kevin will describe conformal prediction approaches to timeseries forecasting using libraries from Nixtla.

"DataCompy - A utility for comparing DataFrames"
speaker: Kevin Kho

In this talk, Kevin will give a tour of Datacompy, an open-source library created by Capital One used for testing. Using Fugue under the hood, Datacompy can handle comparison of Spark, Dask, Ray, Pandas, and Polars DataFrames.

"Unveiling the Future of Autonomous Agriculture: Real-World Applications of the orangutan-stem Project in Data Infrastructure with MLOps"
speaker: Mike Stack (Manager, Data Science & Engineering, 365 Retail Markets)

A presentation on the future of Autonomous Agriculture, MLOps, and data infrastructure. Explore open-source frameworks in a multi-cloud architecture for seamless, low-cost MLOps in real-world agriculture use-cases involving computer vision inference, API source systems, and scientific field surveys for tagging data. The first public unveiling of orangutan-stem since its inception in Fall 2019.

"Open standards for ML model deployment"
speaker: Svetlana Levitan

In this talk we will describe open standards (PMML, PFA, ONNX) that have made model exchange and deployment easier and show how they can be used in some applications.

(filler talk if we have extra time) "How to contribute to LightGBM (and other projects you love)"
speaker: James Lamb

A no-slides, all-live-demo, walkthrough of contributing to LightGBM (https://github.com/microsoft/LightGBM). This will end with actually submitting a pull request, and some discussion of how this could be applied to every machine learning related project.

FAQs:

1. Can I attend this event virtually?

No. This is an in-person event.

We may add a virtual-attendance option or recording of the talks as the date gets closer, but for now those are not guaranteed.

2. How do I sign up to give a talk?

Message James Lamb and Ankush Garg here on Meetup if you'd like to give a 5-10 minute lightning talk.

We love first-time speakers and are happy to help review slides, suggest topics, or answer questions. Any topic of the form "about machine learning and not a sales pitch" is welcome.

3. Will this be too advanced for me?

Absolutely not! All are welcome, and we encourage MLOps enthusiasts of all experience levels to attend.

Acknowledgements

Thanks to mHUB for hosting us in their space!