Kasbi Voice Assistant

AI-Powered Voice Assistant for Ride-Sharing Platform

Kasbi Voice Assistant

About Kasbi Voice Assistant

Kasbi Voice Assistant is a production-grade AI voice assistant integrated into a leading ride-sharing platform. The system uses advanced speech recognition and natural language processing to automate workflows, support users, and enhance the overall user experience for both drivers and passengers.

As AI Team Lead, I spearheaded the development and deployment of this voice AI system, ensuring seamless integration with existing platform infrastructure. The assistant handles various tasks including ride booking, status inquiries, driver-passenger communication, and customer support automation.

Quick Facts

  • Status: Production
  • Year: 2024
  • My Role: AI Team Lead
  • Company: Vesal Gasht
  • Users: 50,000+

Key Features

Speech Recognition

State-of-the-art speech-to-text technology with Persian and English language support for accurate voice commands.

NLP Processing

Advanced natural language understanding to interpret user intent and provide contextually relevant responses.

Workflow Automation

Automated handling of common tasks and queries, reducing support load and improving response times.

Technical Implementation

Core Technologies

  • Custom-trained speech recognition models
  • NLP pipeline for intent classification
  • Real-time audio processing and streaming
  • Multi-turn conversation management
  • Integration with ride-sharing backend APIs

Key Challenges Solved

  • Low-latency response requirements (<500ms)
  • Handling noisy audio environments
  • Multi-lingual support (Persian/English)
  • Scalability for concurrent users
  • Production-grade reliability (99.9% uptime)

My Contributions

  • Led the AI team in designing and developing the voice assistant architecture from scratch
  • Developed speech recognition pipeline optimized for Persian language and ride-sharing context
  • Implemented NLP models for intent classification, entity extraction, and dialogue management
  • Designed integration strategy with existing platform infrastructure and backend services
  • Conducted extensive testing and optimization for production deployment
  • Created monitoring systems for performance tracking and quality assurance

Technology Stack

Speech Recognition NLP Python TensorFlow PyTorch Real-time Processing API Integration Cloud Infrastructure

Impact & Results

50K+

Active Users

40%

Reduction in Support Calls

92%

Accuracy Rate

99.9%

System Uptime

Interested in AI Voice Solutions?

Let's discuss how AI voice technology can transform your product.