Spaces:
Running
Running
Update README.md
Browse files
README.md
CHANGED
|
@@ -1,10 +1,21 @@
|
|
| 1 |
---
|
| 2 |
title: README
|
| 3 |
emoji: 📊
|
| 4 |
-
colorFrom:
|
| 5 |
colorTo: gray
|
| 6 |
sdk: static
|
| 7 |
pinned: false
|
| 8 |
---
|
| 9 |
|
| 10 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
---
|
| 2 |
title: README
|
| 3 |
emoji: 📊
|
| 4 |
+
colorFrom: gray
|
| 5 |
colorTo: gray
|
| 6 |
sdk: static
|
| 7 |
pinned: false
|
| 8 |
---
|
| 9 |
|
| 10 |
+
# YPAI
|
| 11 |
+
|
| 12 |
+
YPAI produces training data and evaluates AI models, agents, RAG, speech and multimodal systems. A network of 50,000 contributors, specialist human review and advanced automation work in one platform, YPAI Collect.
|
| 13 |
+
|
| 14 |
+
**AI Training Data.** Qualified speakers, annotators and domain experts produce speech, multimodal data, demonstrations, preferences and agent tasks, with calibration, independent review and adjudication. Datasets are delivered model-ready, with per-item Data Assurance Records, machine-readable rights, reproducible splits, lineage and signed manifests.
|
| 15 |
+
|
| 16 |
+
**AI Evaluation.** Blind comparisons produce preference judgments and reusable human reference sets, and automated judges are calibrated against human reference judgments. Results are delivered as example-level judgments and structured CSV or JSONL.
|
| 17 |
+
|
| 18 |
+
- Data collection and production: https://ypai.ai/data-collection/
|
| 19 |
+
- Evaluation: https://ypai.ai/data-solutions/evaluation/
|
| 20 |
+
- Available on AWS Marketplace as YPAI AI Training Data and YPAI AI Evaluation
|
| 21 |
+
- Contact: support@ypai.ai
|