Building agents is easy, scaling them for thousands of customers is where the real work hides
Franc0Fernand0 · x · 2026-11-19
The author breaks down how software companies turn Agentic AI prototypes into production products, arguing that getting an agent running is the easy part — serving thousands of customers is where most of the work hides.
Key points:
- Data isolation: every customer's data must stay separate through every retrieval step and every model call.
- Compliance first: regulated-industry customers ask about SOC 2, HIPAA, and ISO 27001 before features.
- Monetization: once it works, you still have to figure out pricing.
He also cites AWS ISV resources like Bedrock AgentCore for deployment and ops, with case studies from PagerDuty, Showpad, and Druva — the latter compressing 30–60 day investigations into minutes-long conversational workflows.
More from coding & agent
- Deepnote rebuilds data apps: build with Codex or Claude, host in Deepnote — alex_verem · 2026-11-19
- NDepend open-sources a C# MCP Server with a complete tutorial and 35-minute video — adnan_hashmi · 2026-11-19
- Creator Studio ships with in-app assistants, MCP support, and open Keynote format — jasonwuishere · 2026-11-19
- Vercel's agent-desktop lets Claude Code drive native Mac apps via the accessibility tree — Arindam_1729 · 2026-11-19
- Anthropic publishes free 33-page guide on turning repetitive prompts into reusable Claude Skills — aftahi_ai · 2026-11-19
- Dev open-sources finance MCP server with OAuth 2.1, PKCE and plan-based tool gating — Noahowshhh · 2026-11-19