readme update

README.md

@@ -100,6 +100,21 @@ widget:
 
     '
 ---
+# Github and Technical Report
+To view our technical report and for access to other model and dataset related scripts visit our github [Github](https://github.com/khoj-ai/timely/tree/main)
+
+# Timely: An Embeddings Model For Temporal Reasoning
+
+At Khoj, we develop open-source personal AI to simplify how people engage with machines. The RAG component in modern AI systems commonly uses an embedding model to retrieve relevant documents for a user query. This retrieved-context enables accurate and personalized responses. 
+
+However, most of these models struggle with temporal reasoning. For instance, if asked "Where was I last summer?", the model would struggle to understand the framing of that question. It requires us to understand the relativity of time (that 2010 is before 2011), and when summer might be (between May - September). 
+
+When we express dates, we often use shorthands like ‘back in June’, ‘on summer break’, and ‘06/15’; all syntaxes that models don’t presently handle well. As such, your embedding model may not find documents with dates within that specific period. This limitation is significant, given the importance of time and date in language and daily life.
+
+To address this problem, we propose **Timely**, a comprehensive pipeline for date-aware dataset generation, model fine-tuning, and benchmarking. Specifically, our goal is to create models that can:
+
+1. Identify natural language dates in queries and documents better
+2. Can handle relative and soft data filters more naturally (e.g. discerning that June is closer than November when talking about Spring).
 
 # SentenceTransformer based on Snowflake/snowflake-arctic-embed-s