Update README.md

README.md

@@ -1,10 +1,43 @@
 ---
-title: README
-emoji: 🏃
-colorFrom: yellow
-colorTo: green
+title: Forge
+emoji: ✳️
+colorFrom: blue
+colorTo: gray
 sdk: static
 pinned: false
+short_description: Data sourcing and licensing for AI teams
 ---
 
-Edit this `README.md` markdown file to author your organization card.
+<img src="https://data.whynotforge.com/forge-wordmark.png" alt="forge" width="180" />
+
+# Good models start with the right data.
+
+Forge helps AI teams source and license datasets for model training and evaluation. We work with data owners and authorised suppliers to find collections that fit a specific use case.
+
+[Explore the marketplace](https://data.whynotforge.com/) · [Meet Forge](https://whynotforge.com/) · [LinkedIn](https://www.linkedin.com/company/forge-webdev/)
+
+---
+
+## Our current focus
+
+**Code repositories** — Private codebases with merged PR/MR history and development context for coding agents, training and evaluation.
+
+**Legacy COBOL** — Codebases for AI-assisted modernisation, training and evaluation, with licensing scope agreed per collection.
+
+**Egocentric video** — First-person activity footage for embodied AI and robot learning.
+
+We also welcome requirements across speech, text, images and other data categories. Availability and permitted use are confirmed for each collection.
+
+## For AI teams
+
+Share the data type, volume, intended use and licensing scope you need. We start with a relevant overview and agree sample access before any full delivery.
+
+## For data owners
+
+If you own data or have authority to license it for AI use, share a catalogue through the marketplace. Paid licensing terms are agreed per collection.
+
+**Contact:** [harshchauhan@whynotforge.com](mailto:harshchauhan@whynotforge.com)
+
+---
+
+We will publish authorised samples and documentation here as they become available.