Agentic RAG: Smarter Conversations
in progress
Sankari Nair
We are working on an elevated chat experience. Ask questions about you saved content by combining time ranges, semantic content understanding, source filters, and automatic synthesis.
Key capabilities:
Find what you saved during any timeframe (“last week”, “this year”, “past 30 days”)
Discover cards on specific topics (“Python cards from this year”, “machine learning content”)
Locate every place a concept is mentioned (“where did I mention Docker last month?”)
Filter results by where they came from (YouTube, GitHub, NYTimes)
Generate summaries and insights across selected content (“summarize what I learned last quarter”, “key themes in my React notes”)
In short:
A unified system that understands time, content, and source, so you can instantly search, filter, and synthesize anything you’ve saved.
Log In
Sankari Nair
Merged in a post:
Make AI more accurate
N
Nicolas Kousmichoff
Otherwise it's useless. I just asked it a simple yet very important question about a youtube transcript and it got it wrong. See attached.
s
schroell
Looking forward to this enhancement and just for reference here my findings to motivate the RAG topic in attachment.
Sankari Nair
marked this post as
in progress
g
guy
Hi, I wished I could have downloaded X (tweets) to Recall.
Sankari Nair
We will soon be improving our chat LLM, it would be great to retest it when we do.