Merge pull request #136 from nhsengland/Minor_addragcontent

Update ds251_RAG.md

docs/our_work/ds251_RAG.md

@@ -5,7 +5,18 @@ origin: ''
 tags: ['NLP','LLM','GENAI']
 ---
 
-## PLACEHOLDER PAGE - CONTENT COMING SOON!
+
+
+![Stop lying to me Chat GPT Haca Poster](https://github.com/user-attachments/assets/52cc653c-db42-42de-a4bd-76c0fd42c83f)
+
+See our RAG Demos and discussions [here](https://github.com/nhsengland/ds_251_RAG)
+
+
+## Headlines: 
+- LLMs produce more relevant and accurate content when given key information in the context window. 
+- All RAG techniques seek to utilise this strength of LLMs by maximising the use of the context window. 
+- Modern LLM systems will involve many RAG techniques, often coming as standard in AI development architectures.
+
 
 [comment]: <> (The below header stops the title from being rendered (as mkdocs adds it to the page from the "title" attribute) - this way we can add it in the main.html, along with the summary.)
-#
+#