Initial large-scale training of foundation models on massive unlabeled corpora to learn general representations; the most compute-intensive stage of the lifecycle. Primary AI lifecycle stage: pre-training.
224 active AI roles across 45 companies in our index reference Pretraining as of today. New postings fell 34% in the last 30 days versus the prior 30 (59 → 39).
The companies with the most active Pretraining listings are: Capital One (25 roles), NVIDIA (22 roles), Meta (19 roles), Microsoft (19 roles), Google (13 roles).
Pretraining primarily belongs to the pre-training stage of the AI lifecycle. In current hiring, Pretraining roles concentrate at: pre-training (65%), post-training (21%).
The sectors with the most active Pretraining hiring are: Big Tech, AI Frontier, Semiconductors.
Initial large-scale training of foundation models on massive unlabeled corpora to learn general representations; the most compute-intensive stage of the lifecycle.
Primary AI lifecycle stage: pre-training.
As of today, 224 active AI roles across 45 companies in our index reference Pretraining. Hiring concentrates at the pre-training (65%) and post-training (21%) stages. Most common sectors: Big Tech, AI Frontier, Semiconductors. New postings fell 34% in the last 30 days versus the prior 30 (59 → 39).
2 AI roles tagged pretraining.
| Company | Title | Sector | AI score | Other tags |
|---|---|---|---|---|
| Merck | Director, Data Science, Foundation Model AI | Pharma | 9 | Fine-tuning · Multimodal |
| Pfizer | Principal ML/AI Cheminformatics Scientist | Pharma | 8 |