AI for Founders with Ryan Estes Enterprise AI for $20 a Month
Mar 31, 2026
Ankit Dheendsa, CTO of Morphos.ai and creator of Green Vectors, a patent-pending vectorization tech. He describes KII, a chat-style RAG search for businesses. He explains massive vector compression and 4x faster queries. He discusses affordable enterprise RAG at $20/month, edge and sensor use cases, and applying the tech across vision, drones, and medical/offline inference.
AI Snips
Chapters
Transcript
Episode notes
Massive Vector Reduction Cuts Cost And Boosts Accuracy
- Morphos.ai's Green Vectors reduces vector DB size up to 99.5% while improving query speed ~4x and boosting search accuracy toward the 99th percentile.
- Project Gutenberg test: 260GB → 1.3GB, 15M vectors → 76k, enabling huge storage and cost savings.
Fix Latency And Hallucinations Or Lose Trust
- Prioritize accuracy and latency in RAG — slow, hallucination-prone search breaks user trust and adoption.
- If queries take 15–30s and still hallucinate, users open other tabs and stop relying on the tool.
Gutenberg Case Study Proves The Compression Claim
- Ankit demonstrated Project Gutenberg compression: 15 million vectors reduced to 76,000, turning 260GB into 1.3GB.
- That case study highlights storage, speed, and quality improvements that make enterprise RAG affordable at $20/month.