Chart takeaway
Strongest confidence right now
No confidence leader yet
The learning engine still needs enough scored evidence before one scope clearly stands out.
Current area · Learning
LearningLearning
Learning engine
See what Fypia is currently learning from published posts and comments, which strategy is winning in each scope, and what action you should take next.
Optimization showroom
This is the highest-signal decision surface in the product: understand what the model learned, what it ignored, and where automation is safe enough to trust.
The learning area should read like an executive strategy room — clear takeaways first, then the evidence and controls beneath.
Run controls
Current live strategy versions
These are the versions Fypia is actively applying right now. Roll back only if the latest learning change feels worse in live use.
Learning engine is waiting. No scope is currently auto-applying changes yet. Start by reviewing the scope control center and recent timeline.
Workspace
Choose what part of learning to inspect
Switch between a high-level overview, automation controls, and deeper analysis instead of scrolling through one long page.
No scopes are available.
Legend
How to read this page
These labels appear throughout the learning dashboard. Use this legend when you need a quick refresher.
Chart takeaway
Biggest recent lift
No impact snapshot yet
Run learning again after more live activity to compare before and after scores.
Chart takeaway
Where automation is getting stuck
No bottleneck detected
The funnel does not currently show a blocked review or policy stage.
Chart takeaway
Latest run throughput
No progress history yet
Once a run records throughput, this card will show whether learning is moving enough items each cycle.
Auto apply
The engine is allowed to push safe micro-tunes live on its own.
Review only
The engine can score winners, but a human must approve before anything changes live.
Paused
The engine still observes results, but this scope is blocked from applying new tuning changes.
Stability / readiness
Stability tells you risk level; readiness tells you how close the scope is to automatic tuning.
Winning patterns
What Fypia is using right now
Read each winner as a live recommendation: what pattern is currently ahead, why the engine trusts it, and what you should actually do with it.
Label cheat sheet
If you see short labels like default, statement, or neutral
Those are internal strategy shorthand names. Use the quick translations on the right to understand what the engine really means.
post
default = baseline post formula
A safe control version the engine falls back to when no newer post framing has clearly won.
first comment
statement = memory statement opener
A short first comment that adds one nostalgic detail without immediately turning it into a question.
reply
neutral = calm conversational reply
A steady, natural-sounding acknowledgment that adds a fresh detail instead of repeating the commenter back to them.
post
Post copy winner
Current winner
No winner yet
What this means
Fypia has not seen enough published post performance yet to promote a post-copy pattern.
How to use it
Keep publishing, importing, and giving manual feedback so the engine can accumulate enough evidence.
first comment
First comment winner
Current winner
No winner yet
What this means
Fypia has not seen enough first-comment outcomes yet to promote an opening-comment pattern.
How to use it
Keep publishing, importing, and giving manual feedback so the engine can accumulate enough evidence.
reply
Reply tone winner
Current winner
No winner yet
What this means
Fypia has not seen enough reply outcomes yet to promote a reply tone pattern.
How to use it
Keep publishing, importing, and giving manual feedback so the engine can accumulate enough evidence.
Learning goals
What the current winner is good at
| Goal | Winner | Score | Samples | Why |
|---|
Losing patterns
What is underperforming right now
| Scope | Failure type | Label | Evidence | Reason | Example |
|---|
Combination learning
Which strategy combinations are working together
These patterns show which post, first-comment, and reply combinations are winning or losing together as a full engagement chain.
| Outcome | Combination | Post | First comment | Reply | Score | Samples | What it means | Actions |
|---|
Recent adjustments
Latest bounded learning changes
| Profile | Field | Change | Reason |
|---|
Strategy audit
Why strategies changed
| Scope | Action | Summary | Reason | When |
|---|
Recent learning runs
Latest run outcomes
| Run | Trigger | Status | Progress | Run health | Summary |
|---|