WORKSHOP
Build Production Ready Voice AI Agent
Date: 12th March 2026 | 11:00 AM to 01:30 PM
Venue: Workshop Room 2, NIMHANS Convention Centre, Bangalore
FEES:
• Rs.90 for Leaders Pass holders
• Rs.904 for Professionals Pass holders
• Rs.1,349 for Knowledge Pass holders
• Rs.1,799 for Community Pass holders
• Rs.1,999 for Expo Pass holders
(Limited seats available)
Prasath Sekar
Product Manager, TeleCMI
- Build a working Voice AI Agent connected to a real phone number
- Understand telephony + AI integration architecture
- Learn how to handle production challenges (latency, call control, scaling, compliance)
- Deploy a real-world ready voice agent system
- This is not a theory session attendees will build and understand the complete working flow.
- Developers or Anyone building conversational AI agents
Module 1 – Voice AI Architecture Fundamentals
- Understanding telephony basics: PSTN and call routing concepts
- How real phone calls are processed in cloud telephony systems
- AI Voice Stack overview:
- Speech-to-Text (STT)
- Large Language Model (LLM)
- Text-to-Speech (TTS)
- Real-time streaming systems vs. non-real-time AI processing
- Designing low-latency conversational loops
Module 2 – Connecting AI to Telephony
- Connecting an AI agent directly to a live phone number
- Real-time audio bridge architecture (Caller :left_right_arrow: Telephony Platform :left_right_arrow: AI Engine)
- Handling inbound AI calls
- Handling outbound AI calls
- Managing call session states and events
Module 3 – Live Agent Build (Hands-On)
- Designing effective LLM instructions for voice agents
- Connecting STT and TTS services
- Setting up real-time audio streaming
- Bridging the AI engine with telephony
- Testing the AI agent with a live phone call
- Observing real-time call logs and events
Module 4 – Production Considerations
- Complete call lifecycle management (ringing, answered, failed, disconnected)
- Latency optimization strategies
- Error handling and fallback flows
- Logging and monitoring production systems
- Compliance considerations (Indian number restrictions, IP-based access control)
- Horizontal scaling for concurrent calls
Module 5 – Advanced Topics (Time Permitting)
- Designing domain-specific AI personas
- Post-call webhook integration
- Call duration
- Recording URLs
- Disposition status
- AI summary and transcripts
- CRM automation using webhook events
- Multi-agent routing strategies
- Production deployment best practices
- Basic knowledge of APIs
- Basic understanding in python
- Laptop with internet connection
Benefits/Takeaways of this workshop for the attendees (What will attendees do after attending the workshop which they were not able to do before attending this)
- Build a real AI voice agent connected to a live phone number
- Deploy voice AI in production (not just demo environments)
- Understand telephony integration challenges
- Avoid common architecture mistakes
- Confidently integrate AI with SIP/WebRTC infrastructure
- Design scalable voice AI systems
About Speakers
Prasath Sekar is a Product Manager at TeleCMI, specializing in CRM telephony SDK & CTI with a strong focus on API and technology integration. He brings expertise in driving product strategy and enhancing communication platforms. Prasath holds an engineering background from GKM College of Engineering and is based in Chennai, Tamil Nadu, India, where he works on building seamless tech solutions that bridge business needs with innovative telecom and software tools.