ML Monthly Meetup - AutoML, MLOps and Graph Machine Learning

[Important update: we have closed all registration, and submitted attendees list to Microsoft for badge printing and checkin.
* If you registered on AICamp website as instructed, you are all set.
* If you RSVP on meetup here but not registered on AICamp, you can contact us, we might be able to help.
* If you did NOT rsvp at all, we won't be able to help.
]

Welcome to our in-person ML monthly meetup (January) at Microsoft NYC. Join us for deep dive tech talks on AI/ML, food/drink, networking with speakers&peers developers, and win lucky draw prizes.

[Important RSVP instructions]

• Attendees are required to register on: https://www.aicamp.ai/event/eventdetails/W2023013014 (Correct name is required for badges and check in. NO walk-ins, NO access without badge)
• Contact us to submit topics and/or sponsor the meetup on venue/food/swags/prizes. https://forms.gle/JkMt91CZRtoJBSFUA
• Join the community on Slack for events chat, speakers office hour and sharing learning, job openings, projects collaboration. join slack

Agenda:

• 5:00pm~5:30pm: Checkin, Food and Networking
• 5:30pm~5:40pm: Welcome/Sponsor intro
• 5:40pm~7:40pm: Tech talks
• 7:40pm~8:00pm: Lucky Draw & Mingle

Tech Talk 1: MLOps with Apache Airflow
Speaker: Julian LaNeve, Benjamin Lampel @Astronomer
Abstract: At the end of the day ML pipelines are just data pipelines for living software. And, as more and more DS/ML teams standardize on Python, Apache Airflow has emerged as the secret ingredient in MLOps. As an open source, Python based workflow manager, Airflow allows data teams to stitch together all the technologies needed in productionizing ML workloads. Out of the box, not only does Airflow have the rich scheduling APIs vast ecosystem of connectors to express even the most complex pipelines, but it's also flexible enough to layer upon additional frameworks. This talk will go through:
- A high level introduction to Airflow
- Various ML architectures used by members of the Airflow community
- A demo of AstroPythonSDK; a new OSS project that makes it easier for data scientists to write production quality pipelines.

Tech Talk 2: Best Practices and Learnings for ML Forecasting
Speaker: Ram Seshadri @Google
Abstract: For the last 12 months, I have been working with multiple Google Cloud customers to build forecasting models to solve different forecasting challenges. This is the distilled wisdom from the field that he would like to share with you all on forecasting. I will also present some best practices and links to resources to help you navigate these challenges. You will learn the following:
- How forecasting differs from classical ML (regression)
- How to set up your forecasting team for success
- Pitfalls you need to avoid in your forecasting projects

Tech Talk 3: Dynamic Graph Learning for Graph Topology Inference
Speaker: Lev Telyatnikov, PhD candidate @Sapienza University of Rome
Abstract: Dynamic Graph Learning for Graph Topology Inference is a method for inferring the topology of a graph, specifically for cases where the connectivity of the graph is unknown. Attendees will learn about the concept, challenges of dynamic graph learning and its application in inferring the topology of a graph. In addition, I will discuss the latest research and developments in the field.