Episode 37: EP 25: BYOC Anywhere
Paul Paliychuk of Zep AI and Omnistrate co-founder Alok Nikhil lay out ten principles for secure Bring Your Own Cloud deployment: keep the data plane in the customer's account, connect without exposing it, swap static credentials for short-lived identity, and make vendor access revocable by design. They walk through the four levels of BYOC, what changes inside air-gapped environments, and why AI products hit the enterprise security bar far earlier in their life than previous generations of SaaS did. The episode closes with a live demo that turns a Helm chart into a BYOC offering with a self-service customer portal.
Special Guests
Alok Nikhil
Co-founder and CTO of Omnistrate; previously at AWS, where he was the original technical author of Aurora Serverless.
Timestamps
00:00Why AI Products Need Flexible Cloud Deployment01:49How Zep AI Builds Context for AI Agents03:05Zep AI's Journey to Scalable BYOC Deployments04:51The 10 Principles of Secure BYOC07:16Zero Trust Between Vendors and Customers11:58How Zep AI Protects Sensitive Customer Data15:32Choosing Deployment Models for Different Enterprises17:21The Four Levels of BYOC Deployment20:47Managing Software in Air-Gapped Environments25:42Why AI Is Making BYOC an Early Requirement29:20Data and Tenant Isolation in Zep AI30:27Graphiti and Zep AI's Enterprise Graph Database32:18From Open Source to Enterprise-Grade Software35:37Live Demo: Turning a Helm Chart Into a BYOC Offering40:18Configuring the Customer's Cloud Deployment42:14How the Deployment Architecture Works45:29Building a Self-Service BYOC Portal48:17Using the Same Process for Other Software Products49:33Final TakeawaysRelated Episodes
Episode 31: Building BYOC for Fortune 500: The Real Story of Enterprise AI Deployment
Daniel Chalef, founder and CEO of Zep AI, walks through the brutal technical reality of deploying AI products into Fortune 500 environments, zero-trust architectures, Service Control Policies, no-egress VPCs, and broken image supply chains. He explains why pure SaaS AI products can't meet enterprise security bars and how Bring Your Own Cloud (BYOC) deployments actually get shipped into locked-down customer infrastructure.
Episode 18: Deploying Agentic Applications: From Multi-Tenant to "SaaS Anywhere"
In today's session, we're unpacking one of the most critical questions for modern AI companies: How do you actually deploy agentic applications? Join us as we break down how agentic systems will be deployed at scale, what changes SaaS companies must prepare for, and what the future of agent-driven architectures looks like.
Episode 36: Removing Data Consistency Bottlenecks : Building High Margin Scalable AI Business on Cloud
Joe McCunney, CEO of Scalar Labs, makes the case that reliable, consistent data across environments is the hidden bottleneck for enterprise AI, especially in large organizations juggling 300+ database types. He explains how ScalarDB acts as a database-agnostic binder that connects multiple data sources and treats them as a single database, and how the company's open source refactoring agent generates architecture docs and migration plans that cut enterprise design time by up to 85%.