Flutter Developer who builds products using real user research · Open to Work

Built From a
Farmer's Question:
"Where Should
We Sell?"

I grew up on the Jharkhand–Odisha border in a farming family. My father grows the crops. My mother handles the selling decisions. Every harvest, the same question came up — where do we sell for a better price? So I started building AgroPredict.

▶ See What I Built ▶Watch Documentary
2
REAL APPS LIVE ON GITHUB
AgroPredict + REPSHIELD — both with downloadable APKs
24h
FEEDBACK → FEATURE
Farmer said "which rice?" — variety selector shipped next day
10+
FARMER CONVERSATIONS
Across Odisha & Jharkhand — still visiting, still learning

AgroPredict is actively being tested and improved. Current focus: calculation accuracy, accessibility, and farmer-driven feature development.

THE PROBLEM
Farmers don't lose money growing crops.
They lose it selling them.

I visited farms across Odisha and Jharkhand. Sat with farmers. Asked where they sell, how much profit they make, whether they know a better price exists nearby.

Almost every farmer sells locally by default — not because it's best. Because they have no easy way to compare.

Every harvest season, the same question came up at home — where should we sell? Different mandis offered different prices, but comparing them was difficult. AgroPredict started as an attempt to make that decision easier for my family and nearby farmers.

FEEDBACK → FEATURE

Every feature has a real origin.

My father in our mango farm
WHERE IT STARTED
My father grows crops on our family farm. Observing harvest and selling decisions helped me identify the market-information challenges AgroPredict aims to address.

Every harvest season, the same question came up at home — where should we sell? Different mandis offered different prices, but comparing them was difficult. AgroPredict started as an attempt to make that decision easier for my family and nearby farmers.

BUILT FROM
✓ Family farming experience — Jharkhand–Odisha border
✓ Farmer conversations across Odisha & Jharkhand
✓ AGMARKNET — Government of India mandi price data
✓ Continuous user feedback — features shipped within 24 hours
WHAT I LEARNED

What Farmers Taught Me

Key insights from conversations — not assumptions, real observations.

💰
Profit matters more than priceFarmers care about take-home after transport — not just the mandi rate on a board.
🌾
Variety matters more than crop name"Rice" is not enough — Basmati, Sona Masuri, Common are completely different decisions.
🔊
Voice beats typingMany farmers prefer to speak the crop name rather than type — especially in Hindi.
📊
Information creates profitKnowing one better mandi exists can change the selling decision entirely.
🤝
Buyer discovery is harder than price discoveryFinding who is buying at a good rate is often the bigger problem.
🏘️
Small farmers lack market connectionsLarge farmers have agents. Small farmers rely on whoever shows up first.
📖 Read Full Field Research Notes →
BUILT WITH FARMER FEEDBACK

Conversations across Jharkhand & Odisha

Farmer Odisha
📍 ODISHA
Farmer Odisha
📍 ODISHA
Farmer Odisha
📍 ODISHA
Farmer Jharkhand
📍 JHARKHAND
Farmer Jharkhand
📍 JHARKHAND
Farmer Jharkhand
📍 JHARKHAND
Farmer Jharkhand
📍 JHARKHAND
BUILDING IN PUBLIC

Everything is Verifiable

All code, research, and progress is publicly available. Nothing to hide.

🎬
Documentary
Real farmers. Real farms. Real problem — filmed across Jharkhand and Odisha.
Watch on YouTube → ⌥
GitHub Repository
Full source code with commit history showing real iterative development.
View on GitHub → 📋
Research Notes
Field research notes from farmer conversations — findings that shaped every feature.
Read Notes →
📱
App Demo
Full app walkthrough — showing the complete farmer journey from crop to decision.
Coming Soon — fixing final screens
DOCUMENTARY + APP DEMO

See the Problem.
See the App.

The documentary shows the real problem — farmers speaking in their own words. The app demo shows what I built to solve it. Both are available publicly.

▶ Watch Documentary

App demo coming soon — currently fixing final screens before publishing.

▶
FARMER DOCUMENTARY · JHARKHAND & ODISHA
APP DEMO COMING SOON
ABOUT ME

Who I Am

Anshu Priya

I'm a self-taught Flutter developer and BCA student from the Jharkhand–Odisha border. I build mobile apps around real problems.

AgroPredict came from watching my parents face the same decision every harvest — my father on the farm, my mother figuring out where to sell. I also built REPSHIELD — a safety app where one shake sends your live GPS location to trusted contacts. No unlock needed.

I taught English for almost a year. Teaching kids who aren't technical taught me the same thing building for farmers did — if someone can use it without asking for help, you got it right.

Built through independent research, direct user conversations, and continuous iteration. Every improvement came from real feedback — not assumptions.

Real problems Field Research User Feedback Driven Real-World Testing Hindi + Voice Self-directed learning
PROJECTS

What I Built

AgroPredict
AGRITECH · FLUTTER · AGMARKNET API · GPS · HINDI + VOICE · FIELD TESTED

A crop price decision-support app that helps farmers decide where and when to sell for better profit. Connects to India's AGMARKNET government data. Full Hindi interface. Built from real farmer feedback across Odisha and Jharkhand.

WHAT IT DOES
GPS MANDINearest mandi with approximate distance and travel time
PROFIT CALCNet profit after estimated transport cost — before the trip
SELL / WAITRule-based advice with plain reasoning — no ML model
VOICE OUTPUTReads advice aloud in Hindi — farmer hears without reading
VARIETIESBasmati vs Common — because farmers think in varieties
HINDI UIFull Hindi interface + ₹/quintal ↔ ₹/kg toggle
CURRENT PROGRESS
✓Interviewed farmers across Odisha and Jharkhand
✓Built working Flutter prototype
✓Integrated AGMARKNET government data
✓Implemented Hindi voice support
✓Added mandi comparison system
✓Added transport profit calculator
WHAT I LEARNED
User feedback is more valuable than assumptions
Simplicity beats complexity for rural users
Profit matters more than price to farmers
Accessibility matters as much as functionality
Voice reduces friction more than any UI improvement
AGMARKNET APIGPS Profit CalculatorHindi UI Voice OutputCrop Varieties Rule-based EnginePrice Date Shown
GitHub → ▶ Documentary 📋 Research Notes V2 Roadmap
REPSHIELD
SAFETY APP · FLUTTER · v1.0.0

A personal safety app for situations where seconds matter. One shake sends your live GPS location to all trusted contacts — no unlock, no delay.

GitHub → Download APK
HOW IT WORKS
SHAKE-TO-SOSNo unlock needed — accelerometer fires instantly
LIVE GPSGoogle Maps link sent via SMS to all contacts
STEALTH MODEDiscreet UI for unsafe situations
SHA-256 LOGTamper-evident incident journal
AGROPREDICT V2

What's Next

Every feature came from real farmer feedback — not a wishlist.

01
Price Date
Show actual data date — not just "today"
DONE ✓
02
Voice Output
Advice read aloud in Hindi on advice screen
DONE ✓
03
Selling Score
🟢 Good · 🟡 Average · 🔴 Wait
BUILDING
04
Shared Transport
Split cost between neighbouring farmers
PLANNED
05
Price History
30-day trend — rising or falling?
PLANNED
06
Offline Mode
Cache prices — works without internet
PLANNED
07
Weather
Rain + perishables = better timing advice
PLANNED
08
Odia Support
Hindi isn't first language for all in Odisha
PLANNED
TECH STACK

Skills & Tools

Flutter
Dart
Python
Geolocator
sensors_plus
speech_to_text
flutter_tts
REST APIs
Firebase
Git & GitHub
Android Build
fl_chart
shared_prefs
Gradle / NDK
State Mgmt
UI/UX Design
CONTACT

Let's Work
Together.

Looking for a Flutter developer role — internship, full-time, or remote. I care about building things that reach people who actually need them. If your team does that, let's talk.

Email Me → GitHub LinkedIn
✉️
EMAIL
priyatech.logic@gmail.com
📱
PHONE
7327950897
📍
LOCATION
Jharkhand–Odisha Border, India
💼
AVAILABILITY
Remote / Hybrid / Onsite — Open Now
🌐
PORTFOLIO
priya-codesdaily.github.io