[Topic]
Strategy & foundations
How models are built and adapted, open-weight versus proprietary, and when owning a tuned model makes sense.
ComparisonBeginner
Open-weight vs proprietary LLMs: what should enterprises choose?
Choose a proprietary API when you need the strongest general model with no infrastructure, and open weights when control, data residency, customization or cost at high volume matter more. Most enterprises end up hybrid: frontier APIs for open-ended work, a tuned open-weight model for narrow, high-volume tasks.
PlaybookIntermediate
How to build domain-specific models from proprietary data
A domain-specific LLM is built by picking one narrow, high-volume, measurable task, turning proprietary data into clean, rights-checked, leak-free examples, and teaching behavior with fine-tuning while keeping changing facts in retrieval. Measure a baseline first and train only if a held-out evaluation shows a real gain.
PlaybookIntermediate
OpenAI fine-tuning is shutting down: how to migrate your fine-tuned models to open-weight models
OpenAI stops new fine-tuning jobs on January 6, 2027, and fine-tuned models on GPT-4o, GPT-4o mini, GPT-4.1 and GPT-4.1 mini stop serving on January 14, 2027. Export your data and evals now, decide whether a newer prompted model is enough, and if not, reproduce the fine-tune with LoRA on an open-weight model and prove parity on a held-out set before you cut over.