phoenix
phoenixPhoenix - Your AI Assistant
Inactive session will be ended in 24 hours
End Session

AI-powered dispatch and customer service assistant

  • Posted:
  • Proposals: 53
  • Remote
  • #4503935
  • OPPORTUNITY
  • Open for Proposals
Khalid H.Safeer H.Awais A.Prince Pratap M.Shoaib A. + 48 others have already sent a proposal.
  • 4

Description

Experience Level: Expert
PROJECT: AI Assistant for TaxiCall Platform

Overview

TaxiCall is a cloud-based taxi booking and dispatch platform written primarily in ASP.NET WebForms (VB.NET) with a Microsoft SQL Server backend.

The objective is to develop an AI Assistant that can access TaxiCall operational data and assist drivers, operators, and customers through natural language interactions.

Initial Requirements

The AI Assistant should be able to:

Query booking information from the SQL database.
Answer questions about customers, bookings, drivers, vehicles, and jobs.
Create and amend bookings through controlled workflows.
Search historical bookings and call records.
Provide business reports and summaries.
Assist operators with dispatch decisions.
Provide customer support through web chat and mobile apps.
Potentially support voice interactions in a later phase.

Current Technology Stack

Frontend:

ASP.NET WebForms
VB.NET
JavaScript
jQuery
Bootstrap
jQuery Mobile

Backend:

Microsoft SQL Server
Stored procedures and direct SQL queries
Existing web services (ASMX)

Mapping / Tracking:

Google Maps APIs
Google Directions APIs
OpenLayers mapping components
Vehicle GPS tracking data

Existing Functionality

The system currently manages:

Customer accounts
Driver accounts
Vehicle details
Immediate bookings
Advance bookings
Dispatching
Route calculations
GPS vehicle tracking
Driver availability
Call recording references
Fare calculations
Customer mobile app
Driver mobile app

Database Access Required

The AI agent will require read access initially to:

Customers
Drivers
Vehicles
Bookings
GPS locations
Call history

Later phases may require write access for:

Creating bookings
Editing bookings
Driver allocation
Customer notifications

Preferred Architecture

Phase 1:

OpenAI API
Python backend
SQL Server connectivity
REST API layer

Phase 2:

Integration into TaxiCall web applications
Customer-facing chatbot
Driver-facing assistant
Voice support

Existing Code Available

Relevant code exists for:

Booking creation
Customer booking screens
Driver allocation
Route mapping
Vehicle tracking
SQL database access

Examples can be provided once the project starts.

Deliverables Required

Proof-of-concept AI assistant connected to SQL Server.
Secure database query layer.
Natural language booking search.
Booking summary and reporting functions.
Documentation and source code.
Deployment guidance.

API credentials, database credentials and server access will be provided after selection of the developer.

New Proposal

Login to your account and send a proposal now to get this project.

Log in

Clarification Board Ask a Question

  • Eight Veer Ltd

    Hey John, should the AI assistant inherit TaxiCall’s existing role-based permissions for operators, drivers, customers and administrators, so every query, report, booking view or future action is restricted by the logged-in user and recorded in an audit trail?

  • CYTNHIA W.

    1. In situations where AI recommendations conflict with dispatcher decisions or real-world operational constraints, should the assistant remain advisory only, or do you eventually envision it taking automated actions?

    2. What would you consider the highest-value use case on day one: booking lookup, reporting, operator productivity, customer support, dispatch assistance, or driver assistance?

    3. Are there existing business rules, stored procedures, or dispatch algorithms that the AI must respect, or will part of the project involve translating operational knowledge into AI-assisted workflows?

    4. How important is auditability? Should every AI-generated answer and recommendation be traceable back to specific records, queries, and data sources for compliance and troubleshooting?

    5. If the project is considered a success six months after launch, what metric would matter most: reduced operator workload, faster booking handling, improved dispatch efficiency, higher customer satisfaction, or reduced support volume?

    John B.8:35pm

    In situations where AI recommendations conflict with dispatcher decisions or real-world operational constraints, should the assistant remain advisory only, or do you eventually envision it taking automated actions?

    I envision it taking automated actions.

    What would you consider the highest-value use case on day one: booking lookup, reporting, operator productivity, customer support, dispatch assistance, or driver assistance?

    Dispatch assistance.

    Are there existing business rules, stored procedures, or dispatch algorithms that the AI must respect, or will part of the project involve translating operational knowledge into AI- assisted workflows?

    Initially existing business rules, then probably AI- assisted workflows.

    How important is auditability? Should every AI-generated answer and recommendation be traceable back to specific records, queries, and data sources for compliance and troubleshooting?

    Yes.

    If the project is considered a success six months after launch, what metric would matter most: reduced operator workload, faster booking handling, improved dispatch efficiency, higher customer satisfaction, or reduced support volume?

    Faster booking handling.