Home / Technology / Ex-OpenAI Dev Unveils Jev AI for Automation
Ex-OpenAI Dev Unveils Jev AI for Automation
21 Sep
Summary
- Jev AI offers faster, cheaper automation decisions.
- It provides calibrated decisions, not conversational text.
- Companies like Vercel saw significant speed improvements.
Hyderabad-based TypeSafe AI has launched Jev, an AI model aiming to transform software automation. Founded by Diogo Almeida, a former OpenAI researcher instrumental in developing ChatGPT, the company addresses the limitations of current AI in automation tasks. Jev operates under TypeSafe AI's System One Model, a new class of frontier AI designed for fast, structured decision-making.
Unlike conversational large language models, Jev generates calibrated decisions with fixed outputs, reportedly making it faster and more cost-effective. The model analyzes background information and user queries simultaneously, bypassing conversational text for direct, structured outputs like Noul (Yes/No percentages), Choice (selected options), and Score (numerical ratings).
TypeSafe AI trained Jev using Reinforcement Learning for Calibrated Decisions, ensuring its confidence levels match accuracy. This allows the AI to act autonomously when confident or escalate to humans when uncertain. The technology has garnered significant developer interest, with companies like Vercel reporting 5 to 18 times faster command safety classification results after switching from OpenAI models to Jev.
