DSAlove
EVERYONE CAN REACH THEIR GOAL · WE FIND THE PATH ·EVERYONE CAN REACH THEIR GOAL · WE FIND THE PATH ·
Layer 1 — Core Loop MVP · Android, closed beta

Everyone can reach their goal. We find the path.

A personal DSA mentor that builds your roadmap, fills your gaps in conversation, and adapts as you learn. No locked content. No “Day 247” guilt mechanics. No grinding for grinding’s sake.

scheduleBeta, soon·blockNo spam ever·See the core looparrow_downward
2 min
Onboarding
~18
Core patterns
0
Locked screens
DSAlove
— The map · The mentor · The bridge
Two pointers·Sliding window·Binary search·Hash map·Stack·Recursion·Backtracking·Trees·Graphs·Heap·Greedy·Dynamic programming·Union-find·Bit manipulation·Trie·Two pointers·Sliding window·Binary search·Hash map·Stack·Recursion·Backtracking·Trees·Graphs·Heap·Greedy·Dynamic programming·Union-find·Bit manipulation·Trie·
— The thesis

Most platforms either lock you out or let you flounder.

Linear curriculums gate content behind progression — you can’t practice what you came to practice. Open problem sets do the opposite — you flounder on topics you don’t have the foundations for. Both treat you like you’re broken.

DSAlove does neither. The student never hears “no.” They hear “yes, and here’s the fastest path.”

— The core loop

Five steps. One promise.

Onboarding to meaningful learning in under three minutes. The loop runs every time you open the app — different inputs, different output.

01
Onboard

Four screens. Two minutes.

Goal · experience · timeline · daily commitment. No typing. No self-assessment. No coding test.

FAANG
3–4 problems/day
3 months
02
Roadmap

The whole map. Visible.

The pattern graph is filtered by your goal, linearized into phases, and shown as a scrollable timeline with a clear finish line.

Phase 1 · Foundations
Phase 2 · Core
Phase 3 · Advanced
flagReady for FAANG
03
Navigate

Tap anything. Anywhere.

Follow the recommendation or jump ahead. If prerequisites aren’t met, the gate builds a bridge to where you wanted to go.

YOU
04
Teach

A mentor that probes.

Socratic conversation. Progressive hints. And when you get stuck — the AI diagnoses whether it’s a missing foundation, and fills it inline.

// mentor
Walk me through what two pointers does on a sorted array.
05
Recalibrate

The roadmap bends.

Every signal feeds the map. Breeze through a pattern, it trims. Struggle, it stretches. Every change is visible — and you can veto any of it.

Sliding window−2 problems
Recursion+3 problems

You hit step 05, the map updates, you tap the next pattern, and the loop starts again. Onboarding → meaningful learning in under three minutes.

— The prerequisite gate

The gate that never says no.

Tap a pattern whose prerequisite isn’t met and the system fires a quick 3–5 question MCQ. Three outcomes — none of them “you can’t be here yet.”

Under two minutes. Card-based. Never feels like a test.

ACE
All — or nearly all — correct

Drop in immediately.

Prerequisite marked validated. The roadmap backfills your chosen topic with fewer, harder problems — no busywork on something you already own.

MID
Some correct · gaps in specific sub-concepts

A targeted micro-bridge.

We diagnose exactly which sub-concepts are shaky and populate problems on those specifically. Then your chosen topic. Path-finding, not remediation.

LOW
Most incorrect · prerequisite isn’t solid

Start at the foundation — framed as the path.

“To get to Sliding Window, we need Two Pointers solid. Let’s start here.” Your destination doesn’t change. We just give you the shortest honest route.

— Teaching

A mentor that asks back — and notices what you’re missing.

Progressive hints

Three levels. Earned, not given.

The AI never spoils. It nudges harder only when nudges aren’t working.

  1. L1
    Conceptual nudge
    Points toward the right mental model
  2. L2
    Structural hint
    Narrows the solution space
  3. L3
    Direct pointer
    Names the approach. You apply it.
Gap detection · the differentiator

When you’re stuck, we don’t just hint harder.

psychology

The AI probes whether the struggle is the current problem or a missing foundation. If it’s the foundation, you get a micro-lesson inline — and then we return to where we were.

Mentor
Before we go on — walk me through what two pointers does on a sorted array.
You
Hmm… one points to the start, the other to the end?
Mentor
Right. And what’s the invariant they maintain as they move? That’s the bit we’re going to use here.
flash_onGap detected · micro-lesson queued inline
— The map

A roadmap that bends. Never breaks.

Every problem you attempt becomes a signal — hints used, time taken, whether you reached the optimal approach, whether a gap was detected and filled. The map reads those signals and adapts.

  • visibilityEvery change is visible. The student sees what changed and why.
  • trending_upFraming is always positive. “You’re stronger here than expected — we’ve trimmed this section.”
  • flagThe finish line never disappears. It may move. It is never hidden.
  • edit_noteYou can veto anything. The AI proposes. The student disposes.
My roadmap
Updated 2 min ago
check
Phase 1
Foundations
Done
timeline
Phase 2
Core patterns
63%
Map updated
+3 problems on Recursion — gap detected during Sliding Window.
lock_outline
Phase 3
Advanced
Soon
flag
Finish line
Ready for FAANG
— Not in the product

Things DSAlove doesn’t do.

We took a thousand “nos” so the product could say one clear yes. Here are the loudest ones.

No 01
Streaks.

Retention mechanic, not a learning mechanic. Out of scope for Layer 1.

No 02
Infinite scroll problem sets.

Curated 3–5 problems per pattern. Enough to validate the loop, not enough to drown in.

No 03
Linear lockout.

Tap any pattern from day one. The gate builds a bridge, never a wall.

No 04
Self-assessment surveys.

“Rate your DSA skill 1–10” is unreliable. We calibrate from actual performance.

No 05
A code runner.

No compiler, no green-tick gaming. The AI critiques the reasoning behind your code.

No 06
Social leaderboards.

Your only comparison is your own past self. Calibration data comes in later, anonymized.

— The build plan

Ship the core. Compound.

Three layers. Each is a self-contained product increment. We ship and prove L1, then deepen, then expand the platform.

Layer 01 · now
rocket_launch

Core Loop MVP.

The whole loop: onboarding, roadmap, free navigation, MCQ gate, Socratic teaching, gap detection, performance-based recalibration.

check Onboarding (4 screens · 2 min)
check Personalized roadmap
check Free navigation + MCQ gate
check Socratic teaching · streaming
check In-conversation gap detection
check Performance-based recalibration
check Curated 3–5 problems per pattern
Ship criteria — end-to-end flow works. First open to meaningful learning in under 3 minutes.
Layer 02 · next
layers

Depth + Calibration.

Objective verification, longitudinal profile, long-term retention. The map becomes a portrait.

add Calibrated tests + post-test analysis
add Student profile (conversation-based)
add Revision + spaced repetition
add Infinite pond (dynamic problems)
add Growth mirror + dashboard
add Goal boundary + pattern locking
add Student overrides (skip / add / reorder)
Ship criteria — test-driven recalibration outperforms performance signals alone.
Layer 03 · later
hub

Platform Expansion.

The teaching engine and roadmap architecture generalize. One platform, multiple interview surfaces.

add AI mock interviewer
add System design learning
arrow_forward Behavioral & communication
arrow_forward Cross-domain progress
Ship criteria — a student can prep DSA + system design through one platform.
Work in progress · Closed beta coming

Be the first to find the path.

We’re building the core loop now. Drop your email and we’ll tell you the moment the beta opens — nothing else, no newsletter, no fluff.

142
On the list
0
Spam emails
1
Email, when it’s live