Contact person: Business Client

Phone:Show

Email:Show

Location: Dubai, United Arab Emirates

Budget: Recommended by industry experts

Time to start: As soon as possible

Project description:
"Project Overview:
I have an existing multi-tenant SaaS system (single database with company_id scoping). Super Admin, Company Admin, and Agent roles are already implemented.
I need a new module added inside this existing project (not a separate app).

The module will include:

1. Chat Widget

Embeddable widget script

Multiple widgets per company

Widget modes: AI Bot only, Live Chat only, Offline form, Hybrid (Bot to Agent)

Custom branding: color, logo, welcome text

Department assignment


2. AI Bot Integration

Connect to LLMs (ChatGPT / DeepSeek or similar)

Company-level bot settings

Model selection

RAG: Upload documents and bot must answer from those documents

Bot to Agent transfer rules

Token usage tracking per model/widget/company


3. Live Chat System (Agent Panel)

Agent inbox: My chats, unassigned, department chats

Real-time chat between user and agent

Internal notes, tags, transfer chat to agent or department

View bot history before transfer


4. Ticketing System

Tickets from offline form

Tickets from bot escalation

Manual ticket creation from chat

Ticket list, filters, priority, status, notes, attachments (if needed)

Ticket timeline


5. Chat & Bot History

Live chat history (search and filters)

Bot conversation history (show transfer reason)

Export transcript (optional)


6. Customer Profiles

Save user details collected by bot (name, email, phone, order ID)

Show customer info to agents

View past chats and tickets


7. AI Model & Token Usage

Select default AI model per company

Token usage per model/widget/company

Basic analytics


The developer must follow existing project structure, UI theme, multi-tenant architecture, auth roles, and coding standards.


---

12-Day Task Plan

Day 1 – Create all UI screens (Admin, Agent, Widget/User) as simple mockups or wireframes matching existing layout.
Day 2 – Implement static UI in the existing project (menus, pages, components, mock data, no backend logic).

Day 3 – Create database tables, models, and base routes for widgets, conversations, messages, tickets, knowledge base docs, token usage, customer profiles (all scoped by company_id).
Day 4 – Build widget CRUD in Admin and generate embed script. Implement working frontend widget shell on an external page.
Day 5 – Implement core live chat: create conversation and enable user-to-agent messaging.
Day 6 – Integrate the first AI model and deliver basic working AI bot replies in the widget. Save token usage.
Day 7 – Add RAG: document upload, indexing, and context-based bot replies per company.
Day 8 – Complete Agent panel: inbox, chat window, internal notes, tags, and transfer functions.
Day 9 – Build full ticketing system: create tickets from offline/bot/agent, ticket list, ticket details, statuses, and priority.
Day 10 – Add histories and user details: live chat history, bot history, and bot user detail form saved to customer profiles.
Day 11 – Add AI model selection, token usage view, and polish UI for Admin and Agent panels.
Day 12 – End-to-end testing, bug fixes, optimization, and handover documentation." (client-provided description)