Simplify org card to minimal progress summary

README.md

@@ -9,33 +9,9 @@ pinned: false
 
 # ASTERIZER
 
-**AI-NATIVE · MODEL TO DEVICE**
+**Building AI — from the model to the device.**
 
-> Building AI — from the model to the device. An end-to-end AI stack: self-hosted large language models, a persistent memory layer, application framework, and the on-device AI that runs them — assembled layer by layer, in-house.
-
-**[Invest in ASTERIZER](https://asterizer.ai/pages/contact.html)** · **[bhargav@asterizer.com](mailto:bhargav@asterizer.com)** · **[Explore the models](https://huggingface.co/collections/ASTERIZER/luna-100m-program-6a9115383c18e52460bf67c9)**
-
----
-
-| AI Products | Owned IP | On-Device AI | LUNA Models |
-|:---:|:---:|:---:|:---:|
-| **4+** | **100%** | **24/7** | **3** |
-
-## The LUNA model family (open weights)
-
-- **[LUNA-100M](https://huggingface.co/ASTERIZER/LUNA-100M)** — pretrain + RAG/MCP SFT
-- **[LUNA-300M](https://huggingface.co/ASTERIZER/LUNA-300M)** — pretrained
-
-## What's built
-
-**Memory Layer** · **Application Layer** · **VOX · Voice AI** · **On-Device & Cloud AI**
-
-## License
-
-All ASTERIZER models, datasets, and tokenizers are released under **CC BY-NC-ND 4.0** — free to use and share, not for commercial sale, with attribution required.
-
-**© 2026 ASTERIZER** · Bengaluru, India · [bhargav@asterizer.com](mailto:bhargav@asterizer.com)
-
----
-
-[asterizer.ai](https://asterizer.ai) · [Products](https://asterizer.ai/pages/products.html) · [Invest / Partner](https://asterizer.ai/pages/contact.html)
+- **LUNA-100M** — pretrained on 4.5B tokens, then RAG + MCP instruction-tuned
+- **LUNA-300M** — pretrained on 4.5B tokens (scale-up from 100M)
+- **LUNA-1B** — in progress
+- **Tokenizer** — multilingual BPE (128K / 64K / 8K)