Public matches show what opponents do. kogni turns your own PRACC imports into round evidence, player patterns, VOD context, and follow-up your roster can act on.
No feature wall. Just the three jobs staff repeat after every practice block.
Run the local parser or import PRACC and compatible JSON files without sending Riot credentials to the backend.
Filter by map, side, economy, opening duel, player, and composition; then land on the matching VOD evidence.
Attach the decision to a player, task, strategy board, opponent plan, or next-block checklist.
Filter the practice record first. Open the exact round second. Use VOD only when the timeline, minimap, economy, or player pattern gives you a reason.
Keep the evidence attached while staff turn it into player feedback, a map plan, an assigned task, or the question the next review needs to answer.
Read player, weapon, economy, opening-duel, composition, and map patterns together. Use the AI Analyst to draft a summary or follow-up without replacing staff judgment.
Move the review into Strategy Board, tasks, goals, schedule, opponent preparation, or a role-scoped team handoff. Keep the source round close to the decision.
kogni is built by VALORANT team managers from Lost Puppies. The public tool is designed for open competitive research; private scrim records remain scoped to the roster that owns them.
Yes. Team search, global and circuit rankings, public team snapshots, compositions, players, matches, and available VOD links are ungated.
Your roster, imported scrims, review notes, tasks, strategy boards, and team operations belong to the private workspace and are not published in Power Rankings.
Entry adds complete review and operations. Pro adds VOD automation, deeper analytics, opponent intelligence, and AI-assisted analysis.
No. Riot session handling stays on the desktop parser. Match payloads can be imported for review, but Riot auth tokens are not stored by the kogni backend.