DEV Community: Dimanjan

Architecting a Sub-25kb Website AI Chatbot for Nepal E-Commerce: Romanized NLP, Gemini Flash & Sub-200ms Latency

Dimanjan — Sun, 20 Sep 2026 08:30:01 +0000

When Nepali businesses integrate standard third-party chat widgets—such as Intercom, Zendesk, or Tidio—they frequently face two critical bottlenecks:

Massive Bundle Bloat: A typical SaaS chat widget introduces 600kb to 1.8MB of compressed JavaScript, firing dozens of third-party analytics trackers, custom fonts, and bloated iframe boundaries. On mobile networks in Nepal (NTC and Ncell 4G), this destroys Google Core Web Vitals, spiking Largest Contentful Paint (LCP) and Interaction to Next Paint (INP).
Failure on Romanized Nepali Slang: Conventional English LLM wrappers stumble when buyers query: > "Dai yo jacket L size ma stock cha ki chaina? Delivery charge New Road ma kati parcha? Voucher code mildaina?"

At Sajedar, we architected a production Website Chatbot Integration Architecture that operates within a strict <25kb client-side footprint while achieving sub-200ms Time-To-First-Token (TTFT) powered by Google Gemini Flash.

Here is an architectural deep dive into how we built it.

1. High-Level System Topology

  [Browser Client]  (< 25kb Vanilla JS Widget)
         │
         │  WebSocket / HTTP POST (Message + Context)
         ▼
  [Cloudflare Edge Gateway / CDN]
         │
         │  TLS 1.3 Termination & DDoS Filtering
         ▼
  [FastAPI Backend Engine (Python 3.11)]
    ├── 1. Romanized Slang Normalizer (Regex + Stemmer)
    ├── 2. Live Catalog Vector / SQL Query (Postgres / WooCommerce REST)
    ├── 3. Gemini 2.5 / 3.x Flash Orchestrator
    └── 4. Real-Time Human Escalation Router (WhatsApp & Slack Webhook)

2. The Client: Zero-Dependency <25kb Embed Script

Rather than injecting heavy React or Vue runtime bundles into customer websites, our client widget is built with vanilla DOM APIs.

/**
 * Lightweight (<25kb) Embed Widget Client
 * Canonical Documentation: https://www.sajedar.com/website-chatbot-integration-nepal
 */
(function(window, document) {
  'use strict';

  const CONFIG = {
    apiEndpoint: window.SAJEDAR_CHAT_API || 'https://api.sajedar.com/v1/chat',
    brandName: window.SAJEDAR_BRAND || 'Sales Assistant',
    primaryColor: '#4f46e5'
  };

  function mountChatWidget() {
    const root = document.createElement('div');
    root.id = 'sajedar-web-chat-root';
    root.innerHTML = `
      <div id="sjdr-chat-bubble" style="position:fixed;bottom:24px;right:24px;z-index:9999;">
        <button id="sjdr-toggle-trigger" style="width:58px;height:58px;border-radius:50%;background:${CONFIG.primaryColor};border:none;color:#fff;cursor:pointer;font-size:24px;box-shadow:0 8px 24px rgba(0,0,0,0.15);">
          💬
        </button>
      </div>
    `;
    document.body.appendChild(root);
  }

  if (document.readyState === 'loading') {
    document.addEventListener('DOMContentLoaded', mountChatWidget);
  } else {
    mountChatWidget();
  }
})(window, document);

Grab the full open-source script on our GitHub Gist.

3. Handling Romanized Nepali Intent & Slang

Nepali e-commerce shoppers do not type formal Devanagari ("नमस्कार! के यो सामान उपलब्ध छ?"). Instead, 98% of queries are Romanized Nepali mixed with English keywords.

We developed an open-source parsing module, nepali-messenger-nlp (NPM package), which deterministicly handles:

Price inquiries: kati parcha, rate k ho, final price kati hola
Stock checks: stock cha ki chaina, available huncha
Delivery zones: Classifying Ring Road vs. Outer Valley delivery surcharges
Deposit intent: Intercepting Rs. 100 advance payment confirmations to slash Cash on Delivery (COD) cancellations from 35% to under 10%.

4. FastAPI & Google Gemini Flash Backend

For enterprise throughput, we pair our tokenization engine with Google Gemini Flash models via FastAPI. Gemini Flash provides exceptional multilingual context fidelity and multimodal vision capabilities (allowing buyers to upload screenshots of clothing or spare parts directly to find matching SKU inventory).

import os
from fastapi import FastAPI
from pydantic import BaseModel
import google.generativeai as genai
import httpx

app = FastAPI(title="Nepal Website Chatbot Engine")

genai.configure(api_key=os.getenv("GEMINI_API_KEY"))

SYSTEM_PROMPT = """
You are the AI Sales and Support Assistant for a premier online retail platform in Nepal.
You understand colloquial Romanized Nepali slang ('kati parcha', 'hajur', 'delivery charge kati?').
If a user is ready to order or requests a human supervisor, trigger the human escalation protocol.
"""

class ChatRequest(BaseModel):
    message: str
    session_id: str

@app.post("/v1/chat")
async def chat_endpoint(payload: ChatRequest):
    model = genai.GenerativeModel(
        model_name="gemini-2.5-flash",
        system_instruction=SYSTEM_PROMPT
    )
    response = model.generate_content(payload.message)

    # Automated WhatsApp / Slack escalation webhook for high-intent shoppers
    if "human" in payload.message.lower() or "urgent" in payload.message.lower():
        await notify_human_team(payload.message, payload.session_id)

    return {"reply": response.text.strip()}

5. Commercial Integration Models in Nepal

Unlike rigid SaaS tiers that charge arbitrary seat fees, website chatbot integration in Nepal requires tailored engineering based on the host codebase (WordPress, Shopify, Next.js, or Laravel):

Setup Fee: Rs. 12,000 to Rs. 100,000, finalized strictly after conducting an exploratory technical audit meeting with the website owner and inspecting the underlying stack.
Execution Rate: From Rs. 5 per active session, ensuring zero recurring waste during seasonal lulls.

For businesses and technical teams planning their website conversational AI rollout, inspect our live deployment and interactive discovery calendar:

👉 Explore the Full Architecture: Website Chatbot Integration Service Nepal

👉 Read our Meta Ads & Messenger Engineering Insights: Sajedar Platform

👉 Explore Open Source Code: Sajedar GitHub

Nepal Payment Gateway Integration Guide: eSewa, Fonepay, Khalti Fees, QR Screenshots & COD Defense

Dimanjan — Fri, 18 Sep 2026 14:17:41 +0000

For any developer or merchant building e-commerce, D2C retail, or SaaS platforms in Nepal, payment gateway integration is the single biggest factor influencing checkout conversion and cash flow velocity.

While Nepal Rastra Bank (NRB) reports over NPR 4 Trillion monthly in retail digital payments, consumer retail remains dominated by a unique dynamic:

Only 1 in 20 online shoppers (~5%) completes payment through an official web redirect payment gateway.
The remaining 95% insist on scanning a static Fonepay/eSewa QR code and sending a screenshot of the payment receipt.

1. Why 95% of Nepali Shoppers Refuse Web Payment Gateways

Daily Muscle Memory: Millions of Nepalis scan QR codes daily at physical grocery shops (kirana), supermarket checkouts (Bhatbhateni), vegetable stalls, ride-sharing (Pathao/InDrive), and restaurants. Scanning a QR code via their mobile banking app is second nature; navigating an unfamiliar web gateway feels alien.
Security Skepticism & Friction: When an online store redirects a shopper to an external gateway page requiring them to type in their eSewa password or wait for an SMS OTP, anxiety spikes. Shoppers fear phishing, wallet deductions without order confirmation, or technical errors, causing 65%+ cart abandonment.

2. The High Expense Barrier of Commercial Gateways

Commercial payment gateway aggregators often charge NPR 15,000 to NPR 35,000 in one-time onboarding fees, alongside annual maintenance charges (AMC) and 2.5% to 3.0% Merchant Discount Rate (MDR).
In addition, they mandate strict corporate KYC: Ward commercial registration, IRD PAN/VAT, and a corporate current account requiring NPR 5,000–25,000 minimum balance. For a new brand selling 5 to 10 items a day, committing NPR 30,000+ upfront before validating market demand is economically unviable.

3. The Bootstrapped Developer Workaround: Automated Gmail API Bank Alert Parsing

To bridge the gap between expensive gateways and tedious manual screenshot checking, developers can implement an automated bank credit alert parsing architecture:

The customer scans a standee Fonepay/eSewa QR code and puts their Order ID in the Remarks field.
The bank (NIC Asia, Nabil, Global IME, eSewa) sends an instant email notification (Credit Alert: NPR XXX received from... Remarks/Ref: ORD-1042).
A background cron worker queries the Gmail API for unread bank alerts.
An automated regex engine extracts the Amount and Remarks to match the Order ID and automatically mark the order as CONFIRMED in under 10 seconds without paying 3% gateway MDR.

Open Source Implementation

We have open-sourced the bank alert parser in nepali-messenger-nlp@1.2.0 on NPM:

import { parseBankCreditAlert, matchPaymentWithOrder } from "nepali-messenger-nlp";

const bankEmailBody = "Dear Customer, Your A/C has been credited by NPR 1,500.00 on 18-Sep-2026. Ref: TXN99824. Remarks: ORD-8842. From RAM SHRESTHA via Fonepay. Current Bal: NPR 42,000.00 - NIC ASIA Bank";

const parsedAlert = parseBankCreditAlert(bankEmailBody);
const match = matchPaymentWithOrder("ORD-8842", 1500, [parsedAlert]);

if (match.isMatched && match.confidence > 0.9) {
  console.log("Payment automatically verified via bank email alert!");
}

4. Important Disclaimers: Temporary Bridge vs. Enterprise Compliance

Automated Gmail API bank alert parsing is a pragmatic, cost-effective bootstrap bridge for early-stage merchants.
As your brand scales past 30 to 50 transactions daily, transitioning to formal, licensed payment gateways (eSewa ePay, Fonepay Merchant API, Khalti) is mandatory for automated refunds, legal audit compliance, and seamless multi-channel scaling.

Read the full engineering guide, HMAC-SHA256 signature samples, and COD defense models at:
👉 Nepal Payment Gateway Integration Guide: eSewa, Fonepay & Khalti

How to Build an Online Clothing Business in Nepal: Sourcing, Sizing & 40% COD Return Defense

Dimanjan — Fri, 18 Sep 2026 13:42:32 +0000

Online apparel is the largest e-commerce vertical in Nepal by order volume, but it also suffers from the highest mortality rate among new founders. Over 60% of online clothing stores close within six months not because of lack of sales, but because of 40% return rates caused by sizing mismatches and trapped working capital.

Here is an operational guide to launching a sustainable fashion brand in Kathmandu, Pokhara, and across Nepal.

1. Sourcing Corridors: CMT vs. India vs. China

Where you manufacture determines your cash flow velocity:

Sourcing Hub	Typical MOQ	Transit Time	Best Used For
Local Kathmandu CMT (Thamel/Kapan)	50 – 100 pcs	7 – 14 days	Oversized tees, hoodies, custom streetwear
Ludhiana / Tirupur (India)	200 – 500 pcs	10 – 18 days	Cotton knits, basics, activewear
Guangzhou / Yiwu (China via 1688)	300 – 1,000 pcs	16 – 25 days	Outerwear, puffer jackets, accessories

2. Solving the #1 Profit Killer: Sizing Mismatches & RTO

Over 65% of customer package refusals at the doorstep happen because the garment does not fit the customer.

How to Reduce Returns from 40% to Under 12%:

Never use generic "S/M/L" labels: Asian sizing differs drastically from European and US sizes.
Standardize exact measurement charts in inches: Chest, waist, shoulder width, and garment length.
Pre-dispatch video confirmation: Sending a 5-second video clip of the garment alongside a measuring tape via Messenger or WhatsApp before courier pickup eliminates 80%+ of "item not as expected" excuses.

3. The Pricing Sweet Spot in Nepal

For direct social commerce (Messenger / TikTok / Instagram), the highest-converting price corridor in Nepal is NPR 1,500 to NPR 2,800. Above NPR 3,000, COD hesitation spikes dramatically; below NPR 1,200, courier fees (NPR 150) and packaging consume the margin.

4. Landed Cost & Margin Modeling

Before placing factory orders, model your landed costs and freight fees using the free Sajedar E-Commerce Profit & Landed Cost Calculator.

Read the full comprehensive playbook at:

👉 How to Start an Online Clothing Business in Nepal: Complete Guide

Profitable Business Ideas in Nepal with Low Investment: Financial Models & Legal Checklist

Dimanjan — Fri, 18 Sep 2026 13:42:15 +0000

Most guides discussing "profitable businesses in Nepal with low investment" offer generic, impractical advice like "open a travel agency" or "start dropshipping luxury watches". In reality, starting a business in Kathmandu, Pokhara, or Terai trade hubs requires understanding local compliance thresholds, working capital traps, and Cash-on-Delivery (COD) failure rates.

Here is an evidence-backed framework of 5 viable businesses with startup capital between NPR 35,000 and 180,000, including unit economics, legal registrations, and payback horizons.

1. Gourmet Condiments & Artisan Hot Sauces (Dalle / Chilli Oil)

Initial Capital Required: NPR 45,000 – 75,000
Gross Unit Margin: 55% – 68%
Break-Even Horizon: 2 to 4 months
Why it works: High shelf-life, non-fragile packaging, high repeat purchase rate, and strong gift culture.
Compliance Must-Know: Any packaged food sold commercially in Nepal legally requires Department of Food Technology and Quality Control (DFTQC) lab certification and food hygiene standards.

2. Micro-SaaS & AI Automation Consulting for Local SMEs

Initial Capital Required: NPR 35,000 – 50,000
Gross Margin: 85% – 92%
Break-Even Horizon: 1 month (first retainers)
The Opportunity: Thousands of retail boutiques across Kathmandu and Pokhara spend 40+ hours per week manually answering "Kati ho price?" in Messenger DMs. Setting up automated customer support flows and order confirmation via Meta Graph API generates immediate monthly retainer income.

3. High-Ticket Niche Sourcing (Specialty Specialty Coffee & Tea)

Initial Capital Required: NPR 90,000 – 160,000
Gross Margin: 42% – 58%
Break-Even Horizon: 3 to 5 months
The Opportunity: Directly sourcing single-origin specialty Arabica from Ilam, Nuwakot, and Sindhupalchok, packaging it under an artisanal brand, and marketing it via TikTok and Instagram.

4. Legal Compliance & Tax Framework in Nepal

Before launching any commercial venture:

Ward Office Commercial Registration: Mandatory within your local ward (वडा कार्यालय).
IRD PAN Registration: Sole proprietorship PAN is free at the Inland Revenue Department (IRD).
VAT Registration Threshold: Mandatory if annual taxable goods turnover exceeds NPR 20 Lakhs (or NPR 50 Lakhs for combined goods & services).

5. Free Sourcing & Unit Margin Planning

To simulate unit costs, China import tariffs, and COD return buffers before risking capital, use the interactive calculator at Sajedar E-Commerce Profit & Landed Cost Calculator.

For full research data on Daraz competition, search volume, and market sizing in Nepal, read the complete study at:

👉 Sajedar Market Intelligence Guide

China-to-Nepal Landed Cost, HS Code Tariffs, and Slashing 35% COD Returns for D2C Brands

Dimanjan — Fri, 18 Sep 2026 13:41:43 +0000

In South Asian cross-border e-commerce—particularly direct-to-consumer (D2C) retail in Nepal—the biggest failure mode isn't ad creative fatigue. It is flawed landed cost mathematics combined with an unmitigated 35% Cash-on-Delivery (COD) return rate.

When importing apparel, consumer hardware, or beauty accessories from Guangzhou, Yiwu, or 1688/Alibaba into Kathmandu, naive importers often calculate unit profit as:

Profit = Selling Price - Factory Cost - Local Courier Fee

This calculation fails in production because it completely omits cross-border freight transit, customs tariffs, assessable VAT, and courier return charges.

1. The True China-to-Nepal Landed Cost Equation

Under Nepal's Customs Act 2064 and the annual Finance Act, importing goods via the Kerung overland corridor or Tribhuvan International Airport (TIA) air cargo requires a strict cascading valuation:

Step 1: Base Assessable Value

Assessable Customs Value = (Factory Price in RMB/USD × Exchange Rate) + Cross-Border Freight

Overland Freight (Kerung/Tatopani): Averages NPR 380 to 450 per kg (14–22 days container transit).
Air Freight (TIA Air Cargo): Averages NPR 1,200 to 1,500 per kg (3–5 days transit).

Step 2: Customs Tariff Assessment

Customs duty is levied directly against the Assessable Customs Value:

10%: Basic raw materials, components, and hardware.
15%: Readymade garments (HS Chapter 61 & 62) and basic footwear.
20%: Cosmetics, skincare, and fashion bags.
30%: High-tier consumer electronics, personal care gadgets, and luxury goods.

Step 3: Value Added Tax (13% VAT)

In Nepal, 13% VAT is not applied only to the base cost—it is compounded across (Assessable Customs Value + Customs Duty).

Total Kathmandu Landed Cost = Assessable Value + Customs Duty + 13% VAT + Border Inward Handling

2. The Cash-on-Delivery (COD) Return Buffer

In Nepal, over 85% of D2C purchases are paid via Cash on Delivery. Outside the Kathmandu Valley (Pokhara, Biratnagar, Butwal, Chitwan), couriers experience delivery failure rates between 18% and 35% due to buyer impulse regret, address unavailability, or delivery delays.

Because couriers charge a reverse shipment penalty (NPR 100–150 per package), an e-commerce brand must reserve at least 18% of their target retail price as an RTO risk buffer. Without this reserve, a seemingly healthy 25% gross margin turns into a net monthly operating loss.

3. Interactive Calculation Tool

To eliminate spreadsheet errors, our engineering team at Sajedar launched an interactive simulation tool:

👉 Launch the Nepal E-Commerce Profit & China Landed Cost Calculator

The simulator allows store owners to:

Toggle between RMB (¥ 1688) and USD ($ Alibaba) sourcing currencies.
Select Overland (Kerung) vs. Air Cargo (TIA) freight corridors.
Automatically apply official HS tariff brackets and 13% VAT.
Calculate net margin after applying the 18% COD return buffer.

4. Validating Product Market Demand Before Capital Deployment

Before locking capital into a 500-unit factory production run, store owners should audit keyword search demand and competitor ad spend in Nepal. You can explore our full methodology at Sajedar Product Market Research.

Architecting Enterprise Facebook Messenger AI Bots: Under 2-Second Webhook Latency & Meta Policy Compliance

Dimanjan — Wed, 16 Sep 2026 06:40:12 +0000

In modern conversational commerce, over 85% of direct customer interactions occur inside Facebook Messenger and Instagram DMs.

However, engineering a high-volume Messenger bot introduces severe systemic engineering bottlenecks:

The 24-Hour Policy Window: Meta strictly limits automated outbound messages to 24 hours after a customer initiates contact.
Romanized Multilingual Slang: Customers type in informal romanized dialects mixed with local phrases ("kati parcha", "delivery charge kasto cha", "L size cha ki chaina").
The 35% COD Cancellation Trap: Impulsive orders placed with zero commitment suffer massive return-to-origin rates.
Foreign SaaS Latency: Daisy-chaining no-code builders (ManyChat, Chatfuel) through translation layers results in 8–15s response latency. In social commerce, every second of latency drops conversion by 12%.

To solve this, our engineering team at Sajedar architected a custom full-stack developer pipeline tailored specifically for sub-2-second conversational sales.

⚡ The Full-Stack Architecture (< 2-Second Response Latency)

Instead of relying on bloated no-code decision trees, we engineer dedicated webhooks directly against Meta Graph API v21.0:

[ Customer DM Event ] 
        │ 
        ▼
[ Meta Webhook & HMAC-SHA256 Verification (X-Hub-Signature-256) ] 
        │ 
        ▼
[ Sub-10ms Slang Normalizer & Entity Extractor (Regex + Fast Tokenizer) ]
        │ 
        ▼
[ Semantic Intent Classification & Redis Conversation Context ]
        │ 
        ▼
[ Deterministic Sales Logic & Meta Graph API Dispatch ] ( < 1,800ms )

🔒 Webhook Signature Verification in Node.js (TypeScript)

To prevent replay attacks and spoofed webhook payloads, all incoming events verify the HMAC-SHA256 signature against your Meta App Secret:

import crypto from 'crypto';
import { Request, Response } from 'express';

export function verifyMetaSignature(req: Request, res: Response, buf: Buffer) {
  const signature = req.headers['x-hub-signature-256'] as string;
  if (!signature) throw new Error("Missing X-Hub-Signature-256");

  const elements = signature.split('=');
  const signatureHash = elements[1];
  const expectedHash = crypto
    .createHmac('sha256', process.env.META_APP_SECRET!)
    .update(buf)
    .digest('hex');

  if (signatureHash !== expectedHash) {
    throw new Error("Invalid signature: Webhook payload tampered.");
  }
}

🛡️ Meta 24-Hour Policy Compliance & Message Tags

Under Meta's Platform Policy, businesses cannot message users after 24 hours without compliant Message Tags. Violating this leads to permanent Facebook Page blocking.

Our production architecture integrates:

POST_PURCHASE_UPDATE: For confirmed tracking numbers, shipping updates, and receipt verification.
CONFIRMED_EVENT_UPDATE: For appointment and delivery slot scheduling.
HUMAN_AGENT Tag: Extends the response window up to 7 days when a certified human technical supervisor responds to customer inquiries.
Recurring Notification Opt-In Tokens: Allowing legal re-engagement for opt-in promotions.

📦 Open-Source NLP Engine: `nepali-messenger-nlp`

To assist engineers building for local South Asian commerce, we open-sourced our core Romanized parsing engine on NPM (DA 94) and PyPI (DA 92):

NPM Package: nepali-messenger-nlp
PyPI Package: nepali-messenger-nlp
GitHub Repository: Dimanjan/sjdrreimagine

import { parseNepaliIntent, classifyNepalLocation } from 'nepali-messenger-nlp';

// Detects Romanized intents in sub-2ms:
const intent = parseNepaliIntent("Dai yo jacket ko price kati ho?");
console.log(intent); // { category: 'PRICE_INQUIRY', confidence: 0.98 }

👨‍💻 Need a Dedicated Messenger Bot Developer?

Rather than paying freelancers $35–$75/hour with zero maintenance guarantees or dealing with fragile no-code SaaS builders, Sajedar's Dedicated Messenger Bot Developers provide:

Transparent Dual Pricing: Predictable volume subscription from NPR 2,500 to NPR 12,000 / month ($20 to $90 USD / mo).
Zero Setup Fees: 48-hour onboarding with custom database and catalog integration.
Assigned Human Technical Supervisor: Daily transcript audits to ensure your bot never hallucinates or leaks sales.

Test our interactive smartphone simulator right in your browser at:

👉 https://www.sajedar.com/messenger-bot-developer

Architecting a 2-Second Facebook Messenger AI Chatbot for Nepal E-Commerce

Dimanjan — Tue, 15 Sep 2026 12:02:56 +0000

In Nepal, more than 85% of social commerce transactions happen not on Shopify or custom storefronts, but directly inside Facebook Messenger and Instagram DMs.

However, running high-volume social commerce in Nepal introduces unique systemic bottlenecks:

The Romanized Nepali Challenge: Customers rarely type in formal Devanagari. They message in Romanized Nepali slang ("kati parcha", "delivery charge kasto cha", "L size cha ki chaina").
The 35% COD Cancellation Trap: Over 85% of deliveries are Cash on Delivery (COD). Without upfront commitment, impulsive buyers cancel packages upon arrival, eroding merchant margins.
Response Time Friction: During peak ad spend hours (7:00 PM – 11:00 PM), response delays exceed 45 minutes, resulting in high abandoned carts.

To solve this, our engineering team at Sajedar built a custom AI sales architecture tailored specifically for the Nepali market.

⚡ The Architecture: Under 2-Second Response Latency

Traditional generic chatbots built on no-code foreign builders (ManyChat, Chatfuel) often take 8 to 15 seconds to round-trip when daisy-chained with translation APIs. In conversational commerce, every second of latency drops conversion by 12%.

Our production pipeline optimizes this to under 2 seconds:

[ Customer DM ] 
       │ 
       ▼
[ Meta Webhook Event ] 
       │ 
       ▼
[ Sub-10ms Nepali Slang Normalizer (Regex & Stemming) ]
       │ 
       ▼
[ Fast Semantic Intent Matcher & Catalog Retrieval ]
       │ 
       ▼
[ Deterministic Sales Logic & Response Dispatch ] ( < 2,000ms )

💰 Slashing COD Return-to-Origin (RTO) with Screenshot Verification

One of the biggest leaks in Nepali e-commerce is the 30–35% package return rate on Cash on Delivery orders. When a customer orders without friction, they often regret the purchase or refuse delivery from courier services like Pathao or Nepal Can Move.

Instead of demanding 100% upfront payment (which kills conversion), we designed an Advance Commitment Flow:

The AI chatbot verifies product availability, color, and size in under 2 seconds.
To confirm the COD booking, the AI requests a symbolic Rs. 100 advance deposit via eSewa or Fonepay dynamic QR.
The customer uploads their payment screenshot directly in Messenger.
The system validates the receipt proof and dispatches the confirmed order.

Result: Slashing cancellation and return rates from 35% down to under 10%, saving thousands in courier return fees every month.

🇳🇵 Open Source NLP Engine

As part of our commitment to the Nepali developer ecosystem, we have open-sourced our core Romanized parsing engine:

NPM Package: nepali-messenger-nlp
GitHub Repository: dimanjan/sjdrreimagine

import { parseNepaliIntent, classifyNepalLocation } from 'nepali-messenger-nlp';

// Detects Romanized intents instantly:
const intent = parseNepaliIntent("Dai yo jacket ko price kati ho?");
console.log(intent); // { intent: 'PRICE_INQUIRY', confidence: 0.94 }

🧑‍💻 The Human Technical Supervisor Advantage

Unlike generic foreign self-serve platforms where you are on your own when a bot hallucinates or breaks, Sajedar's Facebook Messenger AI Chatbot pairs every store with a dedicated human technical supervisor:

Zero technical setup needed from store owners.
Daily transcript auditing to correct edge cases.
Continuous prompt optimization and price sync.
Transparent volume-based pricing from NPR 2,500 to 12,000/month.

🔗 Try the Live Interactive Simulator

Curious how your store would handle high-volume Nepali customer queries?

Test our interactive phone simulator right in your browser at:

👉 https://www.sajedar.com/facebook-messenger-ai-chatbot-nepal