Resources|Podcast|Context Window|EP3Context Window: Enterprise AI Reality Check
with Ed Anuff and Anant Jhingran
Season 1 · Episode 3
Solving the Relevancy Problem: Using Your Data in Models
June 24th, 2025 | 47:03 Runtime
Episode Transcript
TIMESTAMPS
00:00 Introduction and Exciting Developments in AI
02:49 The Quantum Leap: Shifting Perspectives
05:52 Navigating the RAG Landscape
08:42 Understanding RAG: Memory and Data Retrieval
11:55 Challenges in Data Architecture and Retrieval
14:54 The Role of Context in RAG Systems
17:53 Art vs. Science in Prompt Engineering
20:57 The Future of RAG and Context Windows
24:01 Agents and RAG: A Symbiotic Relationship
26:43 The Evolution of AI: From Models to Agents
29:57 Addressing Non-Determinism and Hallucinations
33:08 Conclusion: The Importance of Data Architecture
QUOTES
"The one part that is true is there's a lot of trial and error. There are certain aspects of it that are more art than science."
— Ed Anuff
"If you can't depend on it, it's entertainment."
— Anant Jhingran
"RAG does not equal vector retrieval."
— Ed Anuff
"Even though agents occupy all your attention, a good RAG pipeline is still fundamentally important."
— Anant Jhingran
"Larger context windows don’t eliminate RAG—they just give us more room to optimize."
— Ed Anuff
TIMESTAMPS
00:00 Introduction and Exciting Developments in AI
02:49 The Quantum Leap: Shifting Perspectives
05:52 Navigating the RAG Landscape
08:42 Understanding RAG: Memory and Data Retrieval
11:55 Challenges in Data Architecture and Retrieval
14:54 The Role of Context in RAG Systems
17:53 Art vs. Science in Prompt Engineering
20:57 The Future of RAG and Context Windows
24:01 Agents and RAG: A Symbiotic Relationship
26:43 The Evolution of AI: From Models to Agents
29:57 Addressing Non-Determinism and Hallucinations
33:08 Conclusion: The Importance of Data Architecture
QUOTES
"The one part that is true is there's a lot of trial and error. There are certain aspects of it that are more art than science."
— Ed Anuff
"If you can't depend on it, it's entertainment."
— Anant Jhingran
"RAG does not equal vector retrieval."
— Ed Anuff
"Even though agents occupy all your attention, a good RAG pipeline is still fundamentally important."
— Anant Jhingran
"Larger context windows don’t eliminate RAG—they just give us more room to optimize."
— Ed Anuff