Executive roundtable series

The Agent Scale Circle

Companies are about to have orders of magnitude more AI agents than human employees, and most of today’s context architecture was not built for that scale.

For AI engineering, platform, architecture and technology leaders.

Region
APAC, six cities
Dates
October to November 2026
Access
By invitation
The premise

Built for thousands, not millions.

// per-user assumptions do not survive per-agent traffic

Context architecture designed around human session counts breaks quietly when the callers are software. Agents do not wait, do not batch politely, and do not stop at five o’clock, and the layer between them and your data feels that first.

That is the starting point for this series. Not a product session, and not a briefing. A table of senior people working through what their current latency budget assumes, what happens to it at agent volumes, and which parts of the stack fail first.

Six cities across APAC, between October and November 2026. Each table runs under the Chatham House Rule.

The evidence
10ms Vector search latency after migration, down from two seconds in a published case study
99.5% Improvement in that same latency measurement

Redis customer case study published on redis.io, 2026. Figures pending client verification.

What the table will cover

Three questions on the table.

Each table works through the same three questions, shaped by what the room brings to them.

What your latency budget assumes

The first question is arithmetic: how many calls per task, how many tasks per agent, how many agents. Most budgets were written for a number that is about to be wrong by orders of magnitude.

// tension: the assumption is human-shaped

Cache, memory, or context layer

These three words describe overlapping things and different architectures. This session looks at where each organisation draws the boundaries, and whether the distinction is doing real work or just naming.

// differentiator: an always-current context layer, not a cache

What breaks first at volume

The table works through real failure stories: the component that fell over, whether it was the one anyone predicted, and what the organisation changed afterwards.

// trigger: unified context and memory platforms for production agents
The series

Six cities. One layer.

Each city is its own table with its own room, confirmed independently. Registration of interest is open now, with dates and venues confirmed city by city.

Companies will have orders of magnitude more agents than human beings.

Rowan Trollope CEO, Redis
Before you register

Background reading.

Published Redis material behind the questions on the table. Share your details once to unlock all three.

Platform Redis for agent memory and context Product Vector search and semantic caching Case studies Latency and scale results from production teams
Request an invitation

Join us at the table.

Tell us which city suits you and we will come back to confirm your spot.

Require private transfer?
Who is convening this

About the hosts.

Redis

Redis spent two decades being the thing everyone put in front of their database, which gave it an unusual vantage point: it has always sat exactly where latency is decided.

The company now positions that same position as the context and memory layer for production AI agents, rather than only a cache.

Innovatus Media

Innovatus Media is a Sydney based B2B events agency that runs invitation-only executive roundtable series across APAC, EMEA and North America.

We convene senior decision makers for closed door conversations on the problems they are actually working on, and produce every element of the series end to end, from the room and the table to the follow up.

Questions first

Talk to the team.

Connect for more details on the series.