Three weeks ago I published a 100K-row synthetic sleep health dataset on Kaggle. Here’s what happened:<\/p>\n
– 9,824 views in 20 days<\/p>\n
– 2,158 downloads – 21.9% download rate (1 in 5 visitors downloaded it)<\/p>\n
– 42 upvotes – Silver Medal<\/p>\n
– Stayed above 350 views\/day organically after the launch spike faded<\/p>\n
The dataset has 32 features across sleep architecture, lifestyle, stress, and demographics – and three ML targets: cognitive_performance_score (regression), sleep_disorder_risk (4-class), felt_rested (binary).<\/p>\n
The most shared finding: Lawyers average 5.74 hrs of sleep. Retired people average 8.03 hrs. Your occupation predicts your sleep quality better than your caffeine intake, alcohol habits, or screen time combined.<\/p>\n
Today I released a companion dataset: Mental Health & Burnout in Tech Workers 2026<\/p>\n
100,000 records, 36 columns, covering burnout (PHQ-9, GAD-7, Maslach-based scoring), anxiety, depression, and workplace factors across 12 tech roles, 10 countries, 6 seniority levels.<\/p>\n
The connection to sleep is direct – burnout and sleep deprivation are bidirectionally linked. Workers sleeping under 5 hours average a burnout score of 6.88\/10. Workers sleeping 8+ hours average 3.43. The two datasets share enough overlapping features (occupation, stress, sleep hours) that you can build cross-dataset models or use one to validate findings in the other.<\/p>\n
Key burnout findings:<\/p>\n
– 47.9% of tech workers are High or Severe burnout<\/p>\n
– Managers\/Leads average burnout 7.44 vs Juniors 4.80<\/p>\n
– Remote workers: PHQ-9 depression mean 7.44 vs on-site 5.17<\/p>\n
– Therapy users: PHQ-9 drops from 6.56 \u2192 4.64<\/p>\n
– 73% use AI tools daily – and it correlates with higher anxiety<\/p>\n
Both links in profile. Happy to answer questions about how either was built or calibrated.<\/p>\n<\/div>\n
submitted by \/u\/Mohan137 <\/a>
[link]<\/a><\/span> [comments]<\/a><\/span><\/p>
