Update README.md
Browse files
README.md
CHANGED
|
@@ -241,9 +241,17 @@ export AZURE_OPENAI_API_KEY="your-key-here"
|
|
| 241 |
export AZURE_OPENAI_ENDPOINT="https://your-resource.openai.azure.com/"
|
| 242 |
export AZURE_EMBEDDING_DEPLOYMENT="text-embedding-3-large"
|
| 243 |
```
|
|
|
|
| 244 |
|
| 245 |
-
|
|
|
|
|
|
|
| 246 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 247 |
According to the paper, SalesRLAgent achieves:
|
| 248 |
- **96.7% accuracy** in conversion prediction
|
| 249 |
- Outperforms LLM-only approaches by 34.7%
|
|
|
|
| 241 |
export AZURE_OPENAI_ENDPOINT="https://your-resource.openai.azure.com/"
|
| 242 |
export AZURE_EMBEDDING_DEPLOYMENT="text-embedding-3-large"
|
| 243 |
```
|
| 244 |
+
## Training Data
|
| 245 |
|
| 246 |
+
- Synthetic sales conversations generated using large language models
|
| 247 |
+
- 10,000+ conversation scenarios across different customer types
|
| 248 |
+
- Embeddings captured conversation semantic meaning
|
| 249 |
|
| 250 |
+
## Model Performance
|
| 251 |
+
The model learned to identify key conversation patterns:
|
| 252 |
+
- Technical buyers respond to detailed features
|
| 253 |
+
- Price-conscious customers need ROI justification
|
| 254 |
+
- Early-stage prospects require needs assessment
|
| 255 |
According to the paper, SalesRLAgent achieves:
|
| 256 |
- **96.7% accuracy** in conversion prediction
|
| 257 |
- Outperforms LLM-only approaches by 34.7%
|