---
title: README
emoji: 🪨
colorFrom: gray
colorTo: purple
sdk: static
pinned: false
---

<a href="https://shadowrock.io">
  <picture>
    <source media="(prefers-color-scheme: dark)" srcset="https://a.shadowrock.team/assets/logos/full/shadowrock-logo-white.svg">
    <img alt="ShadowRock" src="https://a.shadowrock.team/assets/logos/full/shadowrock-logo-black.svg" width="380">
  </picture>
</a>

# Where Tech Meets Human Experience

[ShadowRock](https://shadowrock.io) is an AI-specialized systems integrator and Zendesk Premier Partner. We help businesses get real value from their go-to-market technology — CRM, support platforms, and applied AI running in production.

## What we publish here

Community model artifacts from our applied-AI work, validated the way we'd want to find them: pinned base revisions, full quantization manifests, raw benchmark JSON alongside every table, and cards that state exactly what was measured and on which hardware.

Current releases: community quantizations of NVIDIA's Nemotron-3-Embed-8B retrieval model — **FP8** (Ada/Hopper/Blackwell), **NVFP4** (Blackwell-native 4-bit), and **MLX 4-bit** (Apple Silicon).

## Find us

- Web: [shadowrock.io](https://shadowrock.io)
- LinkedIn: [linkedin.com/company/shadowrock](https://www.linkedin.com/company/shadowrock)