Blog

Most AI security discourse fixates on model-level threats. Platform-level threats are a different problem entirely.

AI boosts individual productivity, but without a shared system to organize that work, the gains don’t fully translate to the organization.

This month’s Disclosure covers expanded Projects, new finance-specific Skills and Agents, Matrix workflow upgrades, and improvements to presentation quality, parsing, and model performance.

Fine-tuning an open-source model to match frontier quality is the easy part but serving it cost-effectively is the real challenge.

How Hebbia measures agent quality at scale with a hybrid evaluation methodology.

Hebbia researchers leveraged classic signal processing techniques to build a text detection model smaller than most of the images it classifies.

Seyfarth Shaw expands its partnership with Hebbia to deploy Matrix across its transaction practices to accelerate diligence.

AI made every individual 10x more productive, but no company became 10x more valuable. George Sivulka explains why that’s the most important problem in AI right now.

In February, we shipped features to help you build on shared context and get to insight faster.

After years writing and selling investment research, Paul joined Hebbia as the rare AI platform built to represent that work accurately.

Hebbia announced a partnership with Fitch to bringing credit data directly into Hebbia’s AI platform.

The moat of vertical software comes from deeply understanding and encoding the specific workflows preferences and institutional knowledge of teams and firms which general-purpose AI designed for everyone cannot replicate.