Grab's new tool, Spellvault, allows employees to create large language model (LLM) apps without coding, democratizing access to AI development within the company.
The tool supports a simple three-step process: defining tasks, uploading data to the Knowledge Vault, and specifying APIs for additional data, facilitating user engagement.
Spellvault enhances the relevance of LLM outputs by utilizing retrieval-augmented generation (RAG), which integrates a retrieval step before response generation to improve accuracy.
By combining the capabilities of LLMs with Grab’s extensive internal data, Spellvault aims to create more tailored and effective applications for the company.
Collection
[
|
...
]