PalanerFree on the App Store
Get
From Portals to Protocols: Next-Gen Platform Engineering
WorkFree

From Portals to Protocols: Next-Gen Platform Engineering

Sunday 22 November 2026, 17:30Venue time (Jerusalem)
JFrog - The SwampDirections
FreeTickets sold by luma.com
Book · Freevia luma.com
Add to calendar

Downloads an .ics file · Times are in Asia/Jerusalem · Google Calendar

Join us for an evening of deep technical insights at the Platform Engineering Meetup hosted by JFrog, where industry experts will explore how platform teams are scaling architectures for AI agents, overcoming GPU capacity constraints, and securing internal developer tooling.

Session 1: Designing Internal Developer Platforms for the AI Agent Era

Speaker: Tsofia Tsuriel, Product Manager @ JFrog

Overview: Traditional IDPs were built around a simple assumption: human developers browsing a clean web UI to discover and manage tools. But what happens when the next consumer of your platform is an autonomous AI agent? How do you expose internal capabilities so that human engineers, CI/CD automation, and AI workflows can all consume the same platform without maintaining fragmented tools? Join us as we unpack the critical architectural shifts needed to build an IDP that holds up in the agentic era.

Session 2: Sliding Through the ICE

Speakers: Tomer Elyasim, DevOps Tech Lead & Netanel Kadosh, Senior DevOps @ Kaltura

Overview: Real-time AI avatars need a GPU for live conversations, and the cloud doesn't always have one to give. When the cloud provider answers "no GPUs", what do you do? In this talk, Tomer and Netanel will walk you through how Kaltura got its real-time AI avatars past Insufficient Capacity Errors (ICE). With denser GPUs, any hardware in stock, slot-based autoscaling, and multi-region overflow, we were able to push our platform to support tens of thousands of concurrent sessions.

Session 3: Production-Ready MCP: Unifying Fragmented Tools Under Okta Authentication

Speaker: Nimer Bsoul, Senior DevOps Engineer @ JFrog

Overview: Connecting AI agents to one MCP server is easy; governing dozens of them across an engineering org is not. Without a single gateway, platform teams lose visibility into who or what agent is accessing internal tools, risking unexpected compute costs and unmonitored execution. Learn how JFrog moved from scattered, unmanaged setups to a centralized, production-grade MCP hub integrated directly with Okta for unified authentication, role-based access control, real-time observability, and fine-grained capacity and rate limits.

Ask Palaner

Going to From Portals to Protocols? Ask me anything about it.

I read the organiser's pages and answer in a few seconds.

Answers are AI-generated · Privacy