Product roadmap

Where ReplayLabs is now, and what is coming next.

A public view of the development path, grounded in shipped patch notes and the next product milestones for replay review, team coaching, and integrations.

Current position

Latest shipped

2026-05-19: Free weekly grading unlock and site stats.

Live today

ReplayLabs now combines replay upload, processing, match analysis, dashboard trends, replay groups, teams, billing, weekly free grading, reports, parser compatibility fixes, and API access.

Active focus

The new-subtractor-integration branch is adding a model-assisted 3v3 decision layer that can surface high-confidence coaching moments on match pages.

Next milestone

Boost-aware AI decision review: Boost context is being folded into the 3v3 positioning model so decisions can account for pad routes, recent boost use, and low-boost pressure.

Achievement road

The path from foundation to coaching platform.

Future status labels are directional, not release dates.

Shipped 4

Live 1

In focus 2

Next up 1

Planned 2

Shipped

2026-03-20

Focus areas and public groups

ReplayLabs moved beyond raw match pages with dashboard focus areas, friendlier grading labels, and public replay groups.

Dashboard focus areas identify common or high-impact mistakes
Public replay groups became accessible without an account
Upload progress and grading labels became easier to read

Patch notes - Mar 20, 2026

Shipped

2026-03-26

Payments, replay library, and sessions

Payments went live, replay cards were rebuilt, and the dashboard gained a Last session review block.

Payments launched
My Replays moved to a cleaner grid with gamemode filters
Dashboard Last session, grading visuals, focus-area recency, and real-date trends improved

Patch notes - Mar 26, 2026

Shipped

2026-04-15

2D replay viewer shipped

The 2D Replay Viewer launched so users can watch replay movement on the website without opening Rocket League, then received live timing and score improvements on 2026-05-06.

2D replay playback added to match review
Events timeline integrated so users can jump directly to key moments
Live timestamp, live score, corrected event timing, aligned slider, and start-position fixes followed on 2026-05-06

Patch notes - Apr 15 and May 6, 2026

Shipped

2026-05-19

Free weekly grading unlock and site stats

Free users gained one weekly grading unlock, and the landing page gained a high-visibility public stats section.

Free plan can unlock full grading and premium replay data for one replay per week
Landing page site stats section added
Plans and legal copy updated around the free weekly unlock

Patch notes - May 19, 2026

Live

Live now

Live ReplayLabs product surface

ReplayLabs now combines replay upload, processing, match analysis, dashboard trends, replay groups, teams, billing, weekly free grading, reports, parser compatibility fixes, and API access.

Match review includes grading, event timelines, 2D replay viewing, boost, positioning, demos, bumps, and reports
Dashboard, replay groups, teams, payments, notifications, and free weekly grading unlocks are live
Developer API, Season 23 parser support, group series summaries, and recent parser hotfixes are part of the live product

Current product surface

In focus

AI build

3v3 AI decision-model integration

The new-subtractor-integration branch is adding a model-assisted 3v3 decision layer that can surface high-confidence coaching moments on match pages.

Model scope focuses on challenge, free touch, defensive shadow, support shadow, and rotate-out decisions
Worker inference stores raw predictions in player_ai_decision_events for audit and rollback
API distillation surfaces only recognized high-confidence disagreements as timeline markers, tendency cards, and score adjustments
Initial replacement scope targets noisy no-challenge, failed-rotate-out, and shadowing-failure presentation

In focus

AI build

AI explanations, tendencies, and benchmarks

The same AI work adds explanation traces, user tendency detection, benchmark comparison, and snapshot-aware rollout controls.

Match explanations expose detector, version, context, checks, thresholds, suppression, and optional AI summary data
User tendencies rank repeated patterns with evidence anchors and recommendations
Benchmark comparison can use pro-labeled samples and reports the active AI snapshot context
Dataset and model snapshots support measurable rollout and rollback

Next up

Next AI

Boost-aware AI decision review

Boost context is being folded into the 3v3 positioning model so decisions can account for pad routes, recent boost use, and low-boost pressure.

Decision features include recent boost spent and gained
Nearest small and big pad distance and availability are part of the model context
Surface variants include low-boost overstay, low-boost overcommit, and leaving play for boost too early

Planned

Research

Boost-choice model validation

Offline boost-choice tooling studies whether 3v3 players should stay in play, route through small pads, or leave for big boost before any standalone detector ships.

Research labels pro replay choices as take_small_pad, take_big_pad, or stay_in_play
Manual review must show better results than current missed-boost and boost-overcommit heuristics
Preferred production path remains boost-aware positioning unless standalone boost events validate cleanly

Planned

Later AI

Sequence-based decision model

A later AI phase explores short replay-frame windows instead of single-frame engineered features for stronger decision understanding.

Initial sequence window is planned around recent player, ball, teammate, and opponent movement
3v3 stays the first target mode before mixing in 1v1 or 2v2
The current snapshot-feature model remains the baseline to beat

Follow the build

Shipped work appears in patch notes as it lands.

The roadmap shows direction. Patch notes show the concrete product changes that have made it into ReplayLabs.

View patch notes Upload replay