Glean provides an enterprise AI platform, described as a 'ChatGPT for work', that started by applying transformer models to the historically difficult problem of enterprise search before the generative AI boom.
The company employs a hybrid AI strategy, using proprietary small language models trained on each customer's data for superior semantic retrieval, while leveraging large third-party models (like GPT, Claude) for answer synthesis and reasoning.
A core tenet of Glean's technology is baking user permissions directly into the indexing and retrieval process, ensuring data security and context-aware results.
CEO Arvind Jain's vision is that AI will augment knowledge workers, not replace them, envisioning a future where every employee has a team of AI assistants to achieve a 10x increase in output and business growth.
12 quotes
Concerns Raised
The primary challenge in enterprise AI is the retrieval of correct, up-to-date, and permission-aware information, not model hallucination.
The enterprise search market has a long history of failures, making it a difficult space to innovate in.
Opportunities Identified
Fundamentally changing how knowledge workers operate by providing them with AI assistants and agents.
Using AI to enable a 10x increase in work output and business growth, rather than just for cost reduction.
Shifting the role of software engineers from writing code to designing systems and reviewing AI-generated code.
Creating 'evergreen documentation' and automating complex reporting tasks using AI agents.