marketing_ops · media · workflow
Spotify uses LLMs to deliver personalized recommendation explanations and AI DJ commentary
Spotify's recommendation system relied on cover art and artist familiarity, leaving users without context for why a recommendation might resonate with them personally and limiting discovery of new and niche content.
How it works
Common implementation structure
How this type of workflow is generally built, generalized across documented cases — not tied to any one vendor's stack. Click any stage to read what happens there. Specific products that implement these stages appear in “Tools commonly seen” below.
Stage 1 · User encounters recommendation
A user encounters a music, podcast, or audiobook recommendation on Spotify, traditionally relying only on cover art and artist familiarity.
Tools used
LlamaLlama 3.1 8BvLLMMMLU
Outcome
Users were up to four times more likely to click on recommendations with explanations; fine-tuned Llama models achieved up to 14% improvement on Spotify-specific tasks while significantly reducing costs and latency for AI DJ commentary.
What failed first
Initial zero-shot and few-shot prompting of open-source models revealed challenges with consistent generation style, safety measures, hallucination prevention, and accurate understanding of user preferences.
Results
Time savedsignificantly reduced checkpointing time
Volumeup to four times more likely to click
Cost replacedsignificantly reducing costs and latency
Running since2023
Grounding & classification
Source type: technical build writeup
28 fields verified against source quotes.
content generationconversational aipersonalizationrecommendation systemknowledge baseproduct catalogfailure mode describedhuman review describedmetric backednamed customerproduction runtime claimedtools describedworkflow describedmediasoftwareaccuracy improvementconversion increasecost reductiontechnical build writeupmarketing opsai draft human approval