AI enhanced UX

on YouTube ↗43m

" summary

Machine learning pipelines classify user intent from short keywords to personalize UX, using a proof‑of‑concept built in about a month that ingests CSV dumps (one set was 15 million rows reduced to ~4 million English keywords), auto‑labels with word lists plus GloVe embeddings (GloVe 100), and trains models in a Spark / Kubernetes / Airflow stack (Spark Kubernetes Airflow). Manual test labeling (1,200 queries) yields about 79% accuracy; predictions are served via authenticated API endpoints.

AI-written summary · about

" chapters (17)

1. 0:00Introduction
2. 10:32Predictive marketing campaigns
3. 13:51Predicting hosts preferences
4. 15:20About Morefull
5. 16:41Orchestrator
6. 18:22Pipelines
7. 21:28Dataset
8. 22:28Problem Definition
9. 23:54Auto Labeling
10. 25:10Methods
11. 27:21Models
12. 28:09Results
13. 28:55Predictions
14. 29:45Manual labeling
15. 31:03Adding context
16. 32:03Setup process
17. 32:31QA

" related talks

• 20mEmploying AI to extract actionable insights from qualitative user...Indrasis Mondal, Adam Gorski · Global Artificial Intelligence Virtual Conference 2022
• 31mAI Use Cases: Adding and Maintaining ValueAndrei Witek · Voxxed Days Bucharest 2022
• 40mAI for UX: Using AI to Cut Costs and Time in UXFabian Ziegler · Webinale Berlin 2026
• 38mAI-Powered Conversational InterfacesPaul Tepper · The Artificial Intelligence Conference 2017
• 24mMachine Intelligence gets realBabak Hodjat, Karin Klein, Todd Levy, Drew Conway · DLD New York 2017
• 01h 24mStay Ahead in AI with Product Engineering and AI Native ToolsBasalt Mounir · DSC MENA 2025
• 42mUX+AI: Mastering the UX of AI-Powered ProductsAdi Mazor Kario · Devfest Strasbourg 2024
• 44mHow to Improve UX with Machine Learning: Insights and learnings until nowCarla De Bona · The Conf 2018