AI Pattern Recognition

Know Exactly What
to Fix Next

Focus Areas is an AI-maintained list of your strengths, weaknesses, and areas you're actively improving — derived from every round outcome, confidence rating, and reflection note you log. It updates incrementally as your interview history grows.

How it works

Three steps from interview data to a prioritized improvement plan.

01

Log interviews with confidence ratings

After every round, rate each question 1–5 and optionally write a reflection. These ratings and notes are the raw signals Focus Areas learns from.

02

AI analyzes your delta

When you hit 'Update focus areas,' the system processes only what's changed since last time — new round outcomes, new question ratings, new reflections. No expensive full re-analysis. Just incremental, event-driven updates.

03

A prioritized, living list

You get a ranked list of topics organized by status: Weaknesses to fix, areas Improving, and Strengths confirmed. Each item has an explanation — where the signal came from, why the AI drew that conclusion.

Three signals, one picture

Every Focus Area falls into one of three categories — so you always know where to put your energy.

Weakness

Topics where your confidence is consistently low or where round outcomes show a pattern of struggle. These are your highest-priority prep targets.

Improving

Topics where the AI has seen a positive trend — rising confidence scores, recent positive round outcomes, or encouraging reflection notes. You're on the right track; keep going.

Strength

Topics where your confidence is consistently high and your round outcomes are positive. Confirmed strengths you can lean on and spend less time prepping.

Every conclusion is explained

Focus Areas doesn't just label you — it shows its work. Click any item to see exactly which rounds, questions, and notes contributed to that conclusion, and read the AI's reasoning in plain language.

  • Which specific rounds and outcomes contributed
  • Which confidence ratings were factored in
  • Why a status changed from Weakness → Improving
  • How many data points informed the conclusion
Distributed caching
Weakness

Across 4 system design rounds, you've consistently struggled with cache invalidation strategies and write-through vs. write-behind trade-offs. Confidence ratings averaged 2.1/5. Two round outcomes were 'failed.'

Evidence

Google SWE · System DesignRound failed
Meta E5 · System DesignConfidence: 2/5
Reflection note · Feb 14Note

Incremental by design

As your interview history grows to dozens of rounds and hundreds of questions, you don't want an AI re-reading everything from scratch. Lytmus processes only the changes since your last update — new events, new outcomes, new notes — so updates stay fast no matter how much history you accumulate.

Event queue

Every meaningful action — logging a round outcome, rating a question, writing a reflection — is recorded as an event. Updates process only unread events.

No cold re-reads

Your full history is never re-analyzed from scratch. The AI holds a running picture and patches it with each batch of new signals.

Additive context

Existing Focus Areas are passed as context into each update, so the AI reasons about change and continuity rather than starting over.

Coach integration

Every coaching session is also an event. Insights from your mock interviews feed back into Focus Areas automatically.

Frequently asked questions

What exactly is a Focus Area?

A Focus Area is a specific topic or skill dimension the AI has identified as worth your attention — it could be a consistent weakness (e.g. 'distributed caching'), an active improvement (e.g. 'behavioral storytelling'), or a confirmed strength. Each area has a status, a confidence score, and an explanation of how it was derived.

How does the AI know what my weaknesses are?

Lytmus collects structured signals from your interview history: the confidence scores you rate questions (1–5), the round outcomes (pass/fail), and the reflection notes you write after interviews. When you request an update, the AI analyzes only the changes since the last update — not everything from scratch — and adjusts Focus Areas accordingly.

How is Focus Areas different from just reading my confidence chart?

The confidence chart shows raw averages by topic tag. Focus Areas goes further: it synthesizes across questions, rounds, notes, and coach conversations to produce a narrative understanding — 'you pass DSA rounds but consistently struggle with follow-up questions about time/space complexity' — not just a number.

What does the Improving status mean?

Improving means the AI has observed a positive trend: your confidence on this topic has been rising across recent rounds, or you've had a positive note or coach session. It's not mastered yet, but the trajectory is good. The AI will move it to Strength when confidence is consistently high.

Can I edit or delete a Focus Area the AI created?

Yes. You can edit the description, change the status, add your own Focus Areas manually, or delete any that don't resonate. The AI will incorporate your edits as context in future updates — it won't overwrite changes you've made.

How often should I update my Focus Areas?

After every substantive interview or mock session is the ideal cadence. Practically, updating after each new round logged gives the AI fresh signals to work with. There's a manual 'Update focus areas' button in the UI; future versions will support automatic triggers.

Does Focus Areas feed into the AI Coach?

Yes — every coach session reads your current Focus Areas as context. If your top weakness is 'distributed caching' and your coach session is about system design, the coach will weight distributed caching questions more heavily.

Stop guessing what to prep next

Let your own interview data tell you.

Get started free →