New event: Are you doing anything with LLMs that you think is cool?

[emilieschario]

Type of event:

• [ ] Data Leadership
• [ ] Tool Talk
• [ ] AMAs
• [ ] Panels

Considerations for event:

• Is this panel diverse? Does it represent the LO Community?
• Is this event purposeful? Do we have a clear owner who is focused on community experience?
• Is this scheduled to optimize for time inclusivity (EMEA or APAC friendly)?

Before event:

• [x] Get it on the LO calendar (requires permissions)
  • [x] Event title
  • [x] Event blurb
  • [x] Hosts
  • [x] Zoom link
  • [ ] Form for calendar invites, if you want
• [x] Create discussion Slack channel
• [x] Share in #virtual-events when booked, pointing people to Slack channel and/or form for invites
• [x] 1 week before bump event
• [ ] Send 2-5 questions to panelists a couple of days before
• [x] 1 day before bump event again
• [x] 10 mins before bump event again (can schedule)

After event:

• [ ] Upload to YouTube
• [ ] Share in #announcements
• [ ] Cross link in meetup Slack channel
• [ ] Archive Slack channel

Messaging: :wave: Hey y’all, I’m very excited to have our panel on Thursday. Here is the quick overview:

• :00 we will open the zoom and the goal will be to kick off at the :03
• I will record the session as we’re kicking off
• We will start with brief introductions. I’ll briefly introduce myself + LO. I’ll remind folks about our Slack channel.
• For your intros, please say your name, where you’re based out of, and your current role. Don’t worry about career path now.
• After we do brief intros, I’d remind folks the panel, so here we’ll talk a bit more about your specific career path.
• Sample questions
• Remember: not everyone has to answer every question, so read the room and LISTEN to the other speakers to make sure you’re saying something new.
• Then, we’ll take questions from the audience.
• We will aim to end ON TIME at the :50. But, surely NLT the :55 Thanks and looking forward to it!

[emilieschario]

Integrating LLMs into Data Pipelines:

How have you integrated LLMs into your existing data pipelines? Can you share specific techniques or tools you’ve used to streamline this integration?

Data Preprocessing and Feature Engineering:

What preprocessing steps or feature engineering techniques have you found most effective when preparing data for use with LLMs? How do you handle large datasets to ensure efficient processing?

Improving Model Performance:

What strategies have you employed to improve the performance and accuracy of LLMs in your projects? Are there any specific tuning or optimization methods that have worked particularly well for you?

Handling Unstructured Data:

How do you leverage LLMs to analyze and extract insights from unstructured data (e.g., text, social media posts, logs)? Can you provide examples of how this has added value to your data analysis efforts?

Real-Time Data Analysis:

Have you implemented LLMs for real-time data analysis or streaming applications? If so, what challenges did you face, and how did you address them to ensure low latency and high reliability?

[emilieschario]

@caitlinmoorman jumped in for me. Going to close this.