All workENGINEERING CASE STUDY / 2024 — 2026
FULL STACK / AGRITECH

KisaanMitra.

Start with the farmer’s context.

MY CONTRIBUTION

Founder / team lead / full-stack contributor

PROJECT SCOPE

Public prototype + venture development

TOOLS & TECHNOLOGIES

React / TypeScript / Supabase / Open-Meteo / Pigeon Maps

01 / THE CONTEXT

The problem behind
the product.

KisaanMitra began as an agricultural resource and market-information project and developed into a team venture. The current public edition uses React, TypeScript, and Supabase, following an earlier Next.js/Node.js build. The product brings farm context and multiple agricultural workflows into one application, while the team has presented the venture at YUKTI and BECon.

The challenge

A broad agricultural platform needs a useful starting point. Before showing planning tools and information, it needs to understand the farmer’s location, crop, and immediate need. It also needs to distinguish functioning data integrations from demonstration content while the product is still a prototype.

WHAT I WORKED ON

I founded KisaanMitra and led the team. My public implementation work includes chat-based onboarding and saving farmer-profile updates to Supabase. I supported the team’s YUKTI pitch remotely while Riddhi Rajesh and Panchadip delivered it, and our team later pitched at BECon ’26’s Bengaluru Moonshot round.

02 / ENGINEERING DECISIONS

Where the choices
matter.

01

Collect context before opening the dashboard

Chat-based onboarding collects the farmer’s name, location, crop, and primary need. The implementation writes those values to a user-linked profile, creating a common starting context for the rest of the application. This is an implemented interaction pattern, with no claimed usability-study result.

02

Keep the current product on one managed data layer

The public edition combines a React/TypeScript interface with Supabase authentication and profile persistence. Profile saving updates an existing user record or creates one when needed. This keeps the identity-to-profile flow explicit as the product grows beyond its earlier stack.

03

Separate real integrations from prototype content

The weather widget fetches Open-Meteo data using location coordinates and has a fallback when retrieval fails. The current mandi prices are generated demonstration values, and historical price charts use fixed examples. They demonstrate an interface; they are not verified market feeds.

03 / SYSTEM FLOW

Inside the architecture.

CONCEPTUAL ARCHITECTURE

KisaanMitra / Inside the system

Select a layer. Follow the logic.

Set up a farmer profile

Capture location, crop, and the primary need through the onboarding conversation.

WORKFLOW / 01

An illustrative view of the documented workflow. Positions and signals are conceptual.

Read the complete workflow
  1. 01

    Set up a farmer profile

    Capture location, crop, and the primary need through the onboarding conversation.

  2. 02

    Save the context

    Associate the collected information with the authenticated user’s Supabase farmer profile.

  3. 03

    Explore the relevant tools

    Move into weather, farm planning, crop calendars, resource discovery, and advisory interfaces.

  4. 04

    Develop the venture

    Bring the platform and its product story into team pitches and continue refining the scope beyond the public prototype.

05 / RESULTS & SCOPE

What the evidence
supports.

Top 100

YUKTI innovation finalist · résumé result

BECon ’26

Moonshot pitch · Bengaluru regional round

The team was selected for YUKTI pitching at the IIC Regional Meet on 2 December 2025, from 10,000+ innovations; the Top 100 rank is recorded in my résumé. BECon was a regional pitch milestone. Neither milestone is a claim of farmer adoption, funding, or measured agricultural impact.

Scope & limitations

This is a public prototype with a mixture of integrations and illustrative content. Market prices and several advisory/history panels need verified data before they can support real agricultural decisions. Adoption, yield improvement, and income effects have not been measured here. The next product validation needs reliable information sources and direct evaluation of farmer workflows.

KEEP EXPLORINGCVE Twin

SECURITY / AGENTIC AI