• my talk "Replacing SQL with Kotlin's DataFrame on the Las Vegas Strip" was accepted for KotlinConf 2023

  • gave the speech to a literal standing-room only crowd, thought it went great, got lots of positive feedback

  • a month later, received the offical feedback and review score, which was lower than the average

  • negative comments focused on 2 things: a) live coding font was too small, couldn't read it and b) did not explain the SQL to dataframe code well enough

  • a) was not really under my control, I did not know what screen size I'd have or the layout or size of the room until day-of

  • b) upon reflection, I agreed the commentors who thought my explanations of key concepts were lacking had a point. I had structured my talk to introduce the key concepts, the translation of SQL queries to dataframe functions, via a live coding session -- however, I did rush through this section a bit, plus, was throwing a lot of concepts out all at once

  • I took for granted that the audience would automatically understand the example domain (casino database marketing), instead, that just added some complexity

  • I had compressed about a 55 minute talk into 40 minutes, much of reduction was to the live coding section

  • Although I was open for questions, a big room is far less likely to be interactive and familiar in the way that a smaller, local Meetup is (where I had presented an earlier version of this talk)

  • After reviewing and rewatching, it was pretty clear that I had erred in trying to squeeze too much into a short window -- perhaps some advanced audience members were able to follow along but it probably was not a "beginner friendly" talk

  • Bottom line - finding out my talk was rated below average was highly disappointing and depressing, of course I kept this to myself

  • I felt awful and guilt-ridden, felt I had let down the large audience (of which I was so appreciative that they wanted to hear my talk!), and also the KotlinConf selection committee, the dataframe developers

  • I created an update to my talk where I explained the dataframe equivalent of all the SQL statements I had converted, so people could see in parallel -- but was not ready to admit the low rating, so never published it

  • Had a few ideas for KotlinConf '24, but still felt quite bad about the whole situation and decided to sit this one out and not to submit any talks

  • Recently saw "SQL for Data Scientists in 100 Queries" via Hacker News front page, thought this could be a great way to re-present the concepts... 100 SQL queries is far more than I could ever present in a single talk. Translated the entire document over to Kotlin dataframe (about 88 quesies in total, the rest were examples of interop with other tools, which I filled in with a few dataframe-only concepts and a few plotting examples)

  • The whole point of the initial presentation was to express my enthusiasm for dataframe and its potential, and to describe some of the pros and cons (but mostly pros) I'd experienced when integrating it into a real-world project. I hoped my publicizing my experience and opinions would encourage other developers to try dataframe in their own projects. Once again, I feel terrible that some people felt I did a poor job or didn't meet their expectations -- I get it, it costs a lot to attend and there are multiple presentations in every time slot, so every attendee wants to maximize the content and learning.

  • I sincerely hope that anyone who came away from my talk without a positive impression, or felt it was a poor choice during that time slot, or just wanted more focused SQL to dataframe translations,will see this document and give dataframe (and me) a second chance.

  • To those who gave my presentation a low rating -- I appreciate the honest feedback, I understand your perspective and do not disagree with it -- I will certainly take your comments into consideration while creating any future talks

  • To the KotlinConf selection committee -- I sincerely appreciate that you took a chance on a new presenter, and feel terrible that my talk did not necessarily meet the high standards people expect from KotlinConf (there were numerous excellent presentations). My affinity for Kotlin remains extremely strong and this new project stems from my desire to spread knowledge and enthusaism for dataframe and hopefully widen its reach. I certainly hope that if I submit a presentation for a future KotlinConf, you'll recognize that I've learned a lot through this process and genuinely hope to give a much better presentation in the future.

Thanks for reading! This wasn't easy to write, hope you can empathize.