KisanSathi — Six AI Farm Experts, Keyless and Local

Problem

Farmer-facing AI tools die at the API-key step, answer in English, and hand back generic advice with no live numbers behind it. A smallholder asking when to sell or how to treat yellowing wheat leaves gets a paragraph, not a decision.

Business Impact

India has more smallholder farmers than any advisory service can staff, and the ones who need advice most are furthest from a key, a card, or reliable bandwidth. A keyless multi-agent assistant that answers in Hindi with today's mandi price is deployable on a village kiosk, not just a demo laptop.

System Approach

• Router agent classifies agent, language, place, and commodity in a single fast LLM call
• Six specialist agents defined declaratively in a registry — key, name, emoji, system prompt, optional tools
• Live tools per domain: Open-Meteo 7-day forecast, Agmarknet daily prices via data.gov.in
• FastAPI backend streams answers; Next.js 16 UI shows the pipeline live — expert picked, language detected, tokens arriving
• Adding an agent is one registry entry; router, API, and UI pick it up automatically
• pytest with async mode for the router and tool layer

Key Decisions & Trade-offs

• Keyless by default — most farmer-facing tools die at the API-key step, so that step had to go
• Local Ollama over a hosted model so answers work offline-ish and no farmer data leaves the machine
• Router plus specialists over one big prompt — domain tools only load where they belong
• Answer in the language asked, not English with a translation toggle
• State honest limits in the UI: LLM advice can be wrong, and pesticide doses or big selling calls should be confirmed at the local Krishi Vigyan Kendra

Current Status

Public and open source (MIT) at github.com/rohitguta2432/kisansathi. All six agents live with the router, streaming UI, and both live data tools working keyless; pytest suite green. Built to be forked — translate it, add crops, add agents, put it on a village kiosk.

Roadmap

• Voice input and spoken answers for low-literacy users
• More crops and region-specific pest libraries
• Offline mandi price cache for intermittent connectivity

What I'd Improve Next

• A hosted instance with a small bundled model so no Ollama install is required
• Wider mandi coverage by falling back across neighbouring districts

Explore More

• MyFinancial — Personal Financial Advisor — Financial planning in India is fragmented across banks, insurance, and tax documents.
• PropCheck — AI Property Trust Score for India — Indian property buyers lose lakhs to fraudulent listings on Magicbricks, 99acres, Housing.
• StellarMIND — Chat-to-SQL with RAG — Business users need to query databases without knowing SQL.
• ClinicAI — WhatsApp AI Clinic Assistant — India has 12 lakh+ small clinics running on phone calls and paper diaries.
• MicroItinerary — AI Travel Planner — Travel apps optimize for proximity and ratings.
• SanatanApp — Hindu Devotional App — Devotional users in India juggle 5+ separate apps for Chalisa, Gita, Aarti, Ramayan, and Mahabharat.
• SynFlow — Enterprise Intelligence Platform — Private deal networks rely on manual introductions and spreadsheets.
• FinBaby (Jama) — Personal Finance Tracker — Indian middle-class families track expenses across UPI apps, bank statements, and paper notebooks.
• RetailOS — Multi-Tenant Retail SaaS — Indian kirana stores and small retailers use paper registers or basic billing software with no inventory tracking, no GST compliance, and no offline support.
• TripHive — Offline-First Collaborative Trip Planner — Group trip planning is fragmented across WhatsApp, Google Docs, Maps, Splitwise, and email.
• ScamRakshak — On-Device AI Scam Detector — Indians lose thousands of crores annually to digital scams via WhatsApp, SMS, and social media.
• PaisaGuard — Family Budget Survival App — Middle-class families worldwide track expenses inconsistently — UPI apps show transactions but don't enforce budgets.
• rohitraj.tech — Engineering work is often invisible.
• Agent Autopsy — Forensic Debugger for Failed AI Agent Runs — Most AI agent failures are silent — the run completes, the status code is green, and the result is still wrong.
• tinyvoice — Fine-Tune a Model in Your Own Voice in an Afternoon — Training your own language model sounds like a PhD job that needs a GPU cluster, so most developers never try.
• snap3d — One Photo In, an Editable 3D Model Out — A photo shows you one side of an object; the other five sides are a guess.
• Agentic OS — Force-Directed Map of a 387-Skill Claude Code Setup — An agent setup grows one plugin at a time until it has hundreds of skills, and nobody can see what it actually contains.
• claude-autodev — Autonomous 8-Stage Dev Pipeline for Claude Code — Agentic coding tools stop at a diff.
• MarginChef — AI Agent That Finds a Restaurant's Margin Leaks — Restaurants run on 3-5% net margins while food costs are up roughly 35% since 2019, and most owners never see per-dish economics.
• Quorum — Deep-Research Agent Swarm with a Shared GraphRAG Brain — Most AI research tools are one model in a loop.
• RegexForge — Plain English to a Regex You Can Trust — Regex is not hard because the syntax is exotic.
• Skillet — Turn Any Docs Page Into an Installable Claude Code Skill — Agent skills are the fastest way to teach a coding agent a new tool, but writing a good SKILL.
• Ladle — Open-Source Prep Forecasting and Food-Cost Leak Detection — Restaurants throw away 4-10% of the food they buy before it reaches a plate, and most owners find out at month end from a food-cost percentage that moved the wrong way, with no idea which ingredient did it.
• Casita — Design Your Home in 3D in the Browser — Home design tools want an install, a login, or a CAD background.
• VoxelForge — A Voxel Sandbox Engine Built Properly — Voxel sandboxes are the demo everyone builds with a coding agent right now, and almost all of them are a throwaway single HTML file that hitches on chunk generation, ships a texture pack, and cannot be read or extended.
• prompt-ocean — Type a Sea, Watch It Exist — Generative interfaces usually hand a model the wheel and hope.
• HEXAPOD — Inverse Kinematics and Gait Simulator — Hexapod gait and leg IK are usually explained with equations and a video, or hidden inside a robotics library.
• avatar-sync — Real-Time Face and Hand Tracking in One HTML File — Face and hand tracking demos come wrapped in a build step, a server, and usually an API key — which puts a wall in front of anyone who just wants to see what the models actually output before building on them.
• Reliability & Production Readiness — Load testing, observability, and API contracts.
• Open Source Repos — Browse the source code behind these projects.