Add organization card

README.md

@@ -1,41 +1,35 @@
----
-title: IOTEverythin
-sdk: static
-pinned: false
----
-
 # IOTEverythin
 
-We build compact, production-oriented text-to-speech voices for conversational AI, with a
-focus on Indian-English for customer-support and website voice assistants.
-
-## VozVox
-
-These voices power [VozVox](https://www.vozvox.com/), our voice agent platform for
-AI-powered conversational agents across phone and web, with natural-sounding speech.
+We build **small, trustworthy AI that runs where the data lives** — on your
+own hardware, over your own documents, with no cloud in the loop.
 
-## The Roxi TTS line
+## What we care about
 
-A family of Indian-English voices fine-tuned from the open MOSS-TTS models. Two tiers: the
-0.1B models run in real time for live agents, and the 1.7B model is for pre-rendered or
-premium-quality audio.
+- **Grounded over fluent.** Models should prove their answers, not perform
+  them. Every answer with a receipt; every unknown admitted.
+- **Local-first.** Consumer GPUs and edge devices are enough for
+  verification-grade AI — your documents never leave your machine.
+- **Honest evaluation.** Held-out calibration, published negative results,
+  and limitations sections that pull no punches.
 
-| Model | Base | Best for |
-|---|---|---|
-| roxi-tts-pro | MOSS-TTS-Local 1.7B | Highest quality and intelligibility, offline or premium |
-| roxi-tts-v3.1 | MOSS-TTS-Nano 0.1B | Real-time, current best small voice |
-| roxi-tts-v3 | MOSS-TTS-Nano 0.1B | Earlier single-speaker voice |
-| roxi-tts-v2 | MOSS-TTS-Nano 0.1B | First Roxi release |
-| roxi-tts-v2-onnx | ONNX build of v2 | CPU inference, no transformers dependency |
-| voxi-tts | MOSS-TTS-Nano 0.1B | Original prototype voice |
+## Featured work
 
-## Focus
+### 🔒 [grounded-pointer-qa](https://huggingface.co/IOTEverythin/grounded-pointer-qa)
 
-- Indian-English accent, natural and telephony-aware
-- Small and fast, built on commercially permissive Apache-2.0 base models
-- Single-speaker branded voices for support calls and website assistants
+A question-answering model that **cannot hallucinate by construction**: its
+output layer can only quote verbatim spans from *your* documents, a trained
+abstention head refuses when the answer isn't there, and decoding is fully
+deterministic. Knowledge is hot-swappable — point it at a folder of
+`.txt`/`.md`/`.pdf` files and it answers from those, no retraining.
 
-## Attribution
+- 74.6 EM on SQuAD v2 (held-out, full-retrieval setting)
+- **91.7% answer precision** in "right or silent" mode
+- 125M params — milliseconds per answer on a consumer GPU
 
-Models are built on MOSS-TTS (Apache-2.0). Training data includes the IIT-Madras Indic TTS
-English set, used with the required copyright notice shown on each model card.
+```python
+from modeling_proqa import GroundedQA
+qa = GroundedQA("proqa.pt")
+qa.load_folder("my/documents")
+qa.ask("when does the contract expire?")
+# {'answer': '30 November 2026', 'confidence': 0.98, 'source': '...'}
+```