01
Tell it how you feel
You pick your emotional state — overwhelmed, stuck, grieving, curious, restless — and the intensity. The app doesn't guess; it asks.
A mobile audiobook app that matches readers to AI-narrated book summaries by feeling — not by genre or popularity. You tell the app where you are emotionally, and it finds the book that meets you there. Same inputs, same result. No black box.
I built it because existing apps sort by genre, popularity, or what an algorithm thinks you'll buy. None of that helps when you're grieving, restless, or looking for direction. BYF asks how you feel, why you want to read, and scores every book through a deterministic engine — feeling, intensity, format, shelf affinity. The result is a match, not a recommendation.
You pick your emotional state — overwhelmed, stuck, grieving, curious, restless — and the intensity. The app doesn't guess; it asks.
The engine scores books across feeling, format, language, and shelf. Same inputs produce the same result every time. No opaque model. No hidden ranking.
AI-narrated summaries in Odia, Hindi, and English. When you're ready for the full book, affiliate links take you there.
The difference is transparency. Most apps optimise for sales or engagement. BYF scores for emotional fit — and the scoring is visible.
User: "I'm feeling lost after a breakup."
Response: Shows trending self-help, romance bestsellers, and "people also bought" based on purchase history.
Optimised for sales. Doesn't know what you need — only what others bought.
User: "I'm feeling lost after a breakup."
Response: Identifies emotional state (grief + identity disruption), matches to books scored for processing loss, rebuilding self-worth, and gentle narrative. Returns a ranked match with a 12-minute AI-narrated summary in your language.
Scored for emotional fit. Transparent. Deterministic.
These are not genres. They are emotional states — the actual reasons people reach for a book. The engine maps each feeling to a curated set of titles scored for intensity, tone, and format.
When the mind won't stop replaying. Books that ground the racing loop and return you to the present moment.
When you can't say no. Books that teach the shape of healthy limits without guilt language.
When the inner voice is cruel. Books that rebuild the relationship between you and you.
When discipline breaks and nothing sticks. Books that name the loop and give you one gate to walk through.
When work eats everything. Books that separate identity from output and teach rest without guilt.
When life feels aimless. Books that don't give answers but ask questions worth sitting with.
The matching engine is rule-based and deterministic. Every score is visible. You can disagree with it.
You tell the app your current feeling (e.g. "I feel directionless"), the intensity (light / heavy), and your preferred format (short summary, deep read, narrative, practical).
The engine maps your feeling to one or more emotional shelves. Each book on those shelves already has a pre-scored profile: feeling weight, intensity curve, format fit, language availability.
Books are ranked by total fit score — the sum of feeling alignment, intensity match, format preference, and shelf affinity. Same inputs, same result. Always.
The top match comes with an AI-narrated summary in Odia, Hindi, or English. When you're ready for the full book, affiliate links connect you directly. No middleman, no paywall on discovery.
I start with Odia because no one else does. Hindi because it's the largest reach. English because the content library lives there. Every summary is AI-narrated with voices tuned for warmth and clarity.
Built first because Odia readers are the most underserved. No other audiobook platform starts here.
The largest reach. AI-narrated summaries for readers who think and feel in Hindi.
Where the content library lives. The scoring engine works the same across all three.
Core UI, recommendation engine, and audio playback are built. Affiliate integration and Odia narration are in progress. BYF is the build that earns — so Mārgadarshak can stay free. One product pays for the other.
Read the full case study