Skip to content

farm-gate-price-discovery-board.md

Latest commit

 

History

History
86 lines (66 loc) · 5.79 KB

farm-gate-price-discovery-board.md

File metadata and controls

86 lines (66 loc) · 5.79 KB
Raw
track rapid-prototype
status published

Farm-Gate Price Discovery Board

Category: Agriculture | Market Access (India) Build time: 3–4 weeks (solo developer)

The Problem

India's 146 million farm holdings — 86% of which are small and marginal — face a persistent price information asymmetry. Farmers bring their produce to the local trader and are offered a price with no reference point. They can accept it or take the perishable produce home. Agmarknet publishes wholesale mandi prices, but the farm-gate price — the actual price the farmer receives at their village — is a black box. Middlemen capture 60-65% of the consumer price, and the lack of transparent farm-gate pricing is a primary mechanism.

The farmer knows what they were offered. They don't know what their neighbour 5 km away was offered for the same quality of produce. They don't know if the trader is offering a fair price relative to the nearest mandi. They don't have a way to broadcast their harvest and let multiple agents bid competitively.

Existing price discovery tools (Agmarknet, mandi boards, WhatsApp groups) cover wholesale mandi prices but not the local farm-gate spot market. The need is for a simple, low-tech, non-AI platform where farmers can anonymously share the prices they're being offered and agents can bid on posted harvest lots.

Current Solutions and Limitations

Approach Limitations
Agmarknet portal Mandi wholesale prices only; not farm-gate; requires internet; not actionable for individual negotiation
Local trader offers Single offer, no comparison; farmer has no bargaining leverage
WhatsApp farmer groups Limited reach; moderated by group admin; prices not structured or searchable
Gramhal chatbot Provides mandi prices but 2-3 days old (as reported by farmers themselves); no bidding
Neighbourhood word of mouth Informal, untimely, limited radius
MSP announcements Only covers notified crops (24 crops); most farmers grow vegetables, fruits, pulses not under MSP

What to Build

  • SMS price post: Farmer sends SMS to a shortcode: "5 quintals, brinjal, ₹400/quintal" → their offer price is published on the board
  • Price board by crop and location: Searchable board showing latest farm-gate prices for each crop in each taluka — "brinjal in Satara taluka: range ₹380-450/quintal, median ₹410"
  • Agent bid system: Registered agents can post bids on farmer harvest listings; farmer sees competing offers via SMS
  • Price trend dashboard: Weekly/seasonal trends for major crops by region — "tomato prices in Kolar have dropped 15% this week, consider holding"
  • Quality disclaimer system: Farmers self-grade produce (A/B/C) with optional photo upload; helps agents bid with confidence
  • Subscription alerts: Farmer subscribes to a crop → receives daily SMS: "Today's brinjal range in Satara: ₹380-440 (10 offers)"
  • Mandi price reference: Alongside farm-gate prices, show nearest mandi wholesale price for comparison

Tech Stack

Component Choice Why
Backend Python FastAPI SMS handling, simple CRUD, quick to iterate
Database PostgreSQL Structured price data; spatial queries for taluka-level grouping
SMS Twilio SMS API (₹0.80/message) or MSG91 (Indian, cheaper at ₹0.25/message) Cheaper SMS; phone verification via OTP
Voice IVR Knowlarity / Exotel (optional Phase 2) Voice interface for farmers who can't read SMS
Frontend HTML + Tailwind via Next.js Desktop board for agents; mobile responsive
Hosting Vercel (frontend) + Railway (backend) Free tier both; no server management

Stakeholders

  • 146 million farm holdings (86% small/marginal, < 2 hectares)
  • ~600 million people dependent on agriculture
  • Local traders and commission agents in 6,000+ mandis
  • Farmer Producer Organisations (FPOs) — 10,000+ across India
  • Mandi boards (APMC — Agricultural Produce Market Committees)
  • Kisan Call Centres — potential distribution partner

Why This Is Hackathon-Sized

  • No AI — pure CRUD + SMS aggregation
  • Text-only interface — no images, no maps, no real-time collaboration
  • Simple data model — farmers, crops, prices, locations, agents
  • Proven SMS pattern — mKisan portal, IFFCO's SMS services are reference implementations
  • Clear incentive — farmers get better prices, agents get more supply visibility
  • No hardware — works on any phone with SMS capability (including feature phones)

MVP Timeline

Week Focus Deliverable
Week 1 SMS ingestion + data model Twilio webhook receives farmer messages; parses crop + qty + price; stores in PostgreSQL
Week 2 Price board web interface Searchable web board showing prices by crop, taluka, date; agent registration + offer posting
Week 3 Agent bid + notification Agents bid on farmer listings; farmer notified via SMS with competing offers; farmer accepts
Week 4 Price alerts + mandi reference Subscription alerts; show nearest mandi wholesale price alongside farm-gate median; pilot in 3 talukas

Risks

Risk Mitigation
Low adoption among older farmers Partner with FPOs and Kisan Call Centres for onboarding; SMS is lowest-tech interface
Agents may collude to suppress prices Anonymous posting prevents collusion; farmer sees multiple agent bids
Price data quality (fake/misleading posts) Reputation scoring; phone number verification; moderation of flagged posts
SMS costs at scale Use Indian SMS provider (MSG91: ~₹0.25/message); reverse-billed (farmers send free shortcode, agents pay for subscription)
Only works for farmers who can read SMS Voice IVR as Phase 2; partner with existing farmer helplines for pilot