AssistBotAI Tech Spec

AI lead capture infrastructure built around RAG, integrations, and automation.

This page is the detailed feature-and-surface reference for AssistBotAI. It maps the product stack, delivery channels, dashboard surfaces, and the operational workflow behind visitor capture and sales routing.

RAG-grounded AI replies

Ground every business assistant on curated chips, business context, and live lead-capture steps.

Integration-ready capture

Run the same capture stack across widget chat, embeddable quick forms, landing pages, and paid campaigns.

Automation after intent

Trigger email and WhatsApp sales alerts the moment a visitor completes a lead-ready flow.

Feature Inventory

What the platform is optimized for

AI and RAG core

Business-aware AI assistant bootstrapped per widget key and business slug.

RAG-ready retrieval pipeline using managed knowledge chips and business-specific context.

Lead-signal orchestration before free-form AI replies so intent is captured first.

Form completion gate so the LLM can be introduced after visitor details are secured.

Integration surfaces

Website AI widget for always-on visitor engagement.

Embeddable quick-form script for campaign landing pages, ad traffic, and partner sites.

Shared tracking model across widget sessions, quick forms, and dashboard analytics.

Business-specific configuration so one platform supports many tenants without rework.

Automation layer

Transactional sales alert emails on completed lead submissions.

WhatsApp sales notifications using approved Twilio templates.

Submission-channel separation between chat form captures and quick-form captures.

Automation designed to fail safely so visitor experience is not blocked by downstream delivery issues.

Dashboards and tracking

Business dashboards for sessions, sources, leads, ads performance, visitor analysis, and form submissions.

Source, campaign, location, repeat-visitor, and website-time context attached to captured leads.

Filter-first workflow for date, source, and channel analysis.

Plain lead-capture pipeline preserved alongside richer AI interactions for sales operations clarity.

System Flow

How AssistBotAI moves from visitor to sales action

1. Capture traffic context

Every session begins with widget or form bootstrap, source tracking, visitor identity hints, and channel attribution.

2. Qualify visitor intent

Lead signals and form fields collect the minimum useful business information before a visitor disappears.

3. Trigger automation

Completed submissions are stored, classified by channel, and routed into sales email and WhatsApp alerts.

4. Review and optimize

Dashboards separate visitors, leads, ads performance, and form submissions so the team can improve what converts.

URL And Surface Reference

Key routes and delivery surfaces

Similar to the PocketLens reference page, this keeps the AssistBotAI platform easy to understand across marketing, business access, and backend delivery surfaces.

Public / marketing

/

AssistBotAI public marketing page with product positioning and business-facing value

/feature/tech_spec

Detailed feature, route, integration, and architecture reference

/local-widget-lab

Local sandbox for testing widget flows without touching production businesses

/embed/quick-form

Embeddable quick-form preview route for campaign and landing-page use

/quick-form.js

Drop-in quick-form script for websites, ads, and campaign pages

/widget.js

Core AI widget script for website chat and guided lead capture

Business access

/b/:businessSlug/login

Business user login

/b/:businessSlug/dashboard

Business operations dashboard and lead-capture management

/b/:businessSlug/leads-login

Lead-facing login surface kept separate from the main business login

/access-hub

Internal access hub with known routes and remembered credentials

Platform operations

/super-admin

Superadmin login

/dashboard

Superadmin workspace for business, users, widgets, and platform controls

/api/quick-form/*

Quick-form bootstrap and submission endpoints

/api/widget/*

Widget bootstrap, chips, chat, and tracking endpoints

Why the architecture matters

One capture stack, many channels.

The strongest part of the current platform is not just the chatbot. It is the shared business logic behind chat, forms, alerts, visitor tracking, and dashboards. That lets AssistBotAI support websites, Meta campaigns, Google Ads landing pages, and direct sales workflows without rebuilding the data path every time.

RAG

Use curated business chips and structured content so the assistant answers from grounded context instead of generic text.

AI

Use AI after qualification, so conversations become useful sales assets instead of expensive open-ended chat.

Integration

Reuse the same business form definitions across website widgets, quick-form embeds, dashboards, and paid acquisition funnels.

Automation

Push completed lead context into operational alerts so sales gets the signal immediately, with source and visitor detail intact.