Mapping opportunity

We're already at work.
Here's the map so far.

Not two products — a roadmap. Every opportunity we've heard, rated for viability, with the first build already named. The rest of the board fills in as we sit with more teams.

Where the conversations stand

The starting point.

AI at Riot won't live in one room. It starts where the pain is clearest — with Broadcast — and expands from there as the vision proves itself.

God's Eye

Already live on this site — a working intelligence engine built from the Sao Paulo event's first scrape across Reddit, YouTube, Twitch, VLR and Kick. Preliminary findings, real data.

Running · Validated in the room
Content Intelligence

In the Broadcast conversations, we surfaced the need for a single morning read that tells Broadcast what content performed, where, and why — so the team stops chasing dashboards and starts the day aligned. It lands inside God's Eye.

Inside God's Eye · Sprint scope
Trailer Studio

In the same conversations, we mapped a faster path from rushes to audience-ready cuts: auto-tagging at ingest, then an assembly engine that turns tagged moments into trailers and social cuts.

Workflow mapped · Architecture next
Physical production

Staging, load-in and resource forecasting. Heard second-hand — we haven't seen how planning actually happens.

Open for discovery
Community & events

Creator and summit content. One gathering, hundreds of local moments — but we haven't met the team that owns it.

Open for discovery
Engineering ops

The internal automations nobody demos. Your engineers would rank these better than we can.

Open for discovery
The roadmap, made real

Every opportunity, rated for viability.

Value against effort, dot size for confidence. Click any dot to see the discovery notes — what we already have, what we'd have to build, and the unknowns the next session would settle.

Impact vs. effort
dot size = confidence
Do first
Big bets
Quick wins
Park it
Physical production
Community & events
Engineering ops
← easiereffort to buildharder →
God's EyeConcept
Viability read · Best build on the board — the engine already works; the launch turns it into Riot's system.
What we already have

A working ingestion and narrative engine with live preliminary findings, plus Broadcast's requirements captured first-hand: real-time event mode, creator and co-streamer chat pairing, regional language views.

What we'd have to build

An eight-week launch: scale the source set, real-time event mode, co-streamer chat pairing, editorial controls, and a live event pilot tuned with the Broadcast team.

What discovery settles

The final source list, which co-streamers make the first roster, and how findings slot into the morning-of routine. The event pilot settles this.

The ranked list
score = impact + ease + evidence
On the map, not yet scored
—
Physical production
Staging, load-in and resource forecasting. Heard second-hand — we haven't seen how planning actually happens.
Unscored
—
Community & events
Creator and summit content. One gathering, hundreds of local moments — but we haven't met the team that owns it.
Unscored
—
Engineering ops
The internal automations nobody demos. Your engineers would rank these better than we can.
Unscored

Two rows carry scores because those are the conversations we've actually had. The rest of the map stays unscored on purpose — sitting with those teams is what turns this from our ranking into Riot's.

From the room

You told us what matters.

Six things came up again and again with Broadcast leadership. Each now has a named home — not a promise, a plan.

React while the event is happening

A live pulse the production team can act on mid-event, while it still changes the show.

Sprint scope · real-time event mode
One morning read

What everyone is talking about, where it came from, and the top creator posts from the last twelve to sixteen hours.

Sprint scope · God's Eye
Co-streamers at scale

What a streamer says, paired with how chat reacts — calibrated on one co-streamer first.

Sprint scope · live chat pairing
Regional language views

Broadcast runs in over a dozen languages; regional teams see the heat in their own region.

Sprint scope · priority broadcast languages
Esports vs game conversations

Separate competitive narrative from general game talk — and package the game-side findings for the game team.

Discovery · with the game team
Attribution to your own posts

“We posted a player feature that day” sitting next to the spike it created.

Phase 2 · after the pilot
Where we'd start

One build now.

Build now: God's Eye

We can build this — in fact, it's already running. The engine is live on real data today, naming the week's biggest Valorant stories with traceable sources. The launch turns it into Riot's system: more sources, real-time event mode, co-streamer chat pairing, editorial controls, and a pilot inside a real event. Every other opportunity on this map stands on the data layer it creates.

  • — A working system today, not a mockup
  • — Scoped and priced from live evidence
  • — The data layer the rest of the roadmap stands on

One engagement, one output: a working God's Eye system in your hands. You decide what happens next with it.

How to read this board

Where we're confident, and where we're not.

What we know

God's Eye is already running on real data, and it's the build we're naming first. Its findings are a first read — the event pilot is what hardens them. Trailer Studio is mapped and rated, but priced only from evidence we gather during the God's Eye build.

What's deliberately unscored

Physical production, community & events, engineering ops. We've heard they matter; we haven't sat with those teams, so we're not going to pretend a number.

How it fills in

Discovery meetings put real divisions on the board. Each session turns a placeholder into a scored row — our ranking becomes Riot's.