5Tracks
5Problem Statements
1Real-Time Conversational AI Meets Agentic Workflows 1 Problem Statements

Theme: From Conversations to Everyday Actions


Problem Statement

How can real-time voice AI simplify everyday life by going beyond conversations to help people manage routines, make decisions, coordinate activities, and complete tasks through natural voice interactions?

What Participants Build

A voice-first, agentic AI companion that understands context, remembers preferences, and uses tool calling and external integrations to perform meaningful everyday tasks.

Example Solutions

  • AI Family Assistant: Manages household schedules, grocery lists, reminders, and family activities.
  • Senior Citizen Companion: Provides companionship, manages daily routines, and coordinates caregiver assistance.
  • Personal Productivity Assistant: Organizes tasks, schedules appointments, and automates everyday activities.
  • AI Learning Companion: Delivers personalized tutoring, schedules study sessions, and tracks progress.
  • Interactive Virtual AI Toy/Robot: Engages users through voice, animations, games, learning activities, and connected-device interactions.
ps21
Agentic Workflow Opportunity Example: AI Family Assistant Before: Retrieve schedules, pending tasks, and user preferences. During: Understand daily priorities and coordinate activities through natural conversations. Action: Update calendars, create shopping lists, assign tasks, or initiate service requests via APIs. After: Send reminders, track completion, and manage follow-ups. Technology Focus Agora Conversational AI, Agora RTC and/or Agora Chat, Murf AI Falcon TTS, LLMs, conversational memory, tool calling, external APIs, optional IoT/Physical AI. Expected Deliverable A working voice AI companion demonstrating natural conversations, contextual awareness, and at least one complete everyday-life workflow involving meaningful actions through external integrations. Note: Participants are encouraged to explore innovative use cases beyond these examples, particularly solutions that combine personalization, real-world automation, and seamless end-to-end user experiences.
2Voice AI at Work: Beyond the Call Center 1 Problem Statements

Theme: From Workplace Assistance to Intelligent Action

Problem Statement

How can real-time voice AI improve workplace productivity by enabling employees to access information, resolve issues, automate repetitive tasks, and complete business workflows through natural, hands-free conversations?

What Participants Build

A voice-first, agentic workplace assistant that combines enterprise knowledge, contextual intelligence, and tool calling to execute real-world business tasks across connected systems.

Example Solutions

  • Field Technician Copilot: Guides troubleshooting, retrieves equipment history, orders spare parts, and updates service tickets.
  • Factory Floor Assistant: Monitors equipment information, assists with operational issues, and triggers maintenance workflows.
  • Warehouse & Logistics Assistant: Checks inventory, tracks shipments, initiates stock transfers, and updates ERP systems.
  • Employee Productivity Assistant: Manages meetings, retrieves company information, creates tasks, and automates routine workflows.
  • AI Training & Onboarding Agent: Delivers interactive training, evaluates employee progress, and updates learning records.
ps31
Agentic Workflow Opportunity Example: Field Technician Copilot Before: Retrieve assigned tickets, equipment history, and maintenance records. During: Diagnose issues through interactive voice troubleshooting. Action: Check spare-part availability, raise orders, update tickets, or escalate incidents through APIs. After: Generate service reports, update records, and initiate follow-ups. Technology Focus Agora Conversational AI, Agora RTC and/or Agora Chat, Murf AI Falcon TTS, LLMs, RAG, tool calling, MCP-compatible integrations, enterprise APIs. Required Demo Flow Participants should demonstrate at least one complete, end-to-end workplace workflow: Initiate: Employee interacts with the AI assistant through natural voice using Agora Conversational AI and Murf AI Falcon TTS. Understand: AI interprets the request, retrieves relevant enterprise information, and identifies the required action. Execute: AI uses tool calling or external APIs to perform meaningful actions, such as updating tickets, checking inventory, or scheduling tasks. Confirm: AI communicates the outcome, handles errors where applicable, and seeks user confirmation for critical actions. Follow Up: AI updates relevant systems, generates a summary, or initiates follow-up activities. Expected Deliverable A working voice AI assistant demonstrating natural conversations, enterprise knowledge integration, and at least one complete workplace workflow involving meaningful actions across connected systems. Note: Participants are encouraged to explore innovative workplace use cases beyond these examples, particularly solutions that demonstrate intelligent automation, multi-step workflows, and measurable improvements in operational efficiency.
3AI Agents That Get Things Done 1 Problem Statements

Theme: From Conversations to Autonomous Action

Problem Statement

How can real-time conversational AI go beyond answering questions to independently plan, coordinate, and execute complex, multi-step tasks across connected applications and services?

What Participants Build

A voice-first, agentic AI assistant that understands user goals, makes contextual decisions, and orchestrates multiple tools and APIs to complete real-world tasks while keeping users informed and in control.

Example Solutions

  • AI Travel Concierge: Plans itineraries, compares travel options, coordinates bookings, and manages changes.
  • Smart Appointment Agent: Finds available slots, schedules appointments, sends confirmations, and handles rescheduling.
  • AI Customer Resolution Agent: Investigates issues, processes eligible requests, updates tickets, and coordinates escalations.
  • Personal Automation Agent: Coordinates calendars, emails, reminders, and tasks across multiple applications.
  • AI Event Planner: Identifies venues, coordinates vendors, manages invitations, and tracks event arrangements.

 

ps51
Agentic Workflow Opportunity Example: AI Travel Concierge Before: Retrieve travel preferences, budget, dates, and existing commitments. During: Research options, compare itineraries, and refine recommendations through natural conversations. Action: Check availability, initiate approved bookings, update calendars, and send confirmations using APIs. After: Track itinerary changes, manage reminders, and coordinate rescheduling when required. Technology Focus Agora Conversational AI, Agora RTC, Murf AI Falcon TTS, LLMs, tool calling, MCP, multi-agent orchestration, external APIs. Required Demo Flow Participants should demonstrate at least one complete, multi-step agentic workflow: Initiate: User communicates a goal through real-time voice using Agora Conversational AI and Murf AI Falcon TTS. Plan: AI interprets the request, identifies dependencies, and determines the required sequence of actions. Execute: AI invokes multiple tools or APIs to retrieve information, coordinate services, and perform tasks. Adapt: AI handles changing requirements, errors, or unavailable options and requests user approval where necessary. Confirm & Follow Up: AI verifies task completion, communicates outcomes, and initiates relevant follow-up activities. Expected Deliverable A working voice AI agent demonstrating natural conversations, multi-step reasoning, at least two coordinated tool/API calls, and successful completion of a real-world task. Note: Participants are encouraged to explore complex workflows, multi-agent collaboration, dynamic decision-making, human-in-the-loop approvals, and intelligent recovery from failed actions. The emphasis is on building agents that deliver verifiable outcomes, not merely conversational responses.
4VibeCart: Shop by Talking 1 Problem Statements

Theme: From Product Discovery to Conversational Commerce

Problem Statement

How can real-time voice AI transform shopping into a seamless conversational experience, enabling customers to discover, compare, purchase, and manage products through natural voice interactions?

What Participants Build

A voice-first, agentic shopping assistant that understands customer preferences, delivers personalized recommendations, and uses tool calling to complete shopping transactions and post-purchase workflows.

Example Solutions

  • AI Personal Shopper: Recommends products based on preferences, budget, and purchase history.
  • Multilingual Retail Assistant: Helps customers discover products, check availability, and place orders in their preferred language.
  • Smart Product Advisor: Compares products, specifications, pricing, and offers to support purchasing decisions.
  • Conversational Checkout Agent: Manages shopping carts, applies discounts, and initiates secure checkout.
  • Post-Purchase Assistant: Tracks orders, initiates returns, and handles customer support requests.
ps41
Agentic Workflow Opportunity Example: End-to-End Conversational Shopping Before: Retrieve customer preferences, purchase history, and available offers. During: Discover, compare, and recommend products through natural conversations. Action: Check inventory, apply discounts, update carts, and initiate checkout through APIs. After: Confirm orders, track deliveries, and assist with returns or support. Technology Focus Agora Conversational AI, Agora RTC and/or Agora Chat, Murf AI Falcon TTS, LLMs, product catalog APIs, inventory integrations, tool calling, optional sandbox payment integrations. Required Demo Flow Participants should demonstrate at least one complete conversational shopping journey: Initiate: Customer engages with the shopping assistant through real-time voice using Agora Conversational AI and Murf AI Falcon TTS. Discover: AI understands customer requirements and retrieves relevant products through catalog APIs. Recommend: AI compares options, checks availability, and provides personalized recommendations. Execute: AI performs meaningful actions through tool calling, such as adding products to a cart, applying discounts, or initiating checkout. Confirm & Follow Up: AI confirms the completed action and provides an order summary, tracking information, or relevant follow-up assistance. Expected Deliverable A working conversational commerce prototype demonstrating real-time voice interaction, personalized product discovery, and at least one successfully executed shopping action through external APIs. Note: Participants are encouraged to explore innovative retail experiences beyond these examples, including multimodal shopping, multilingual interactions, personalized recommendations, and complete pre-purchase and post-purchase journeys.
5Voice AI for Bharat: Speak Your Language 1 Problem Statements

Theme: Breaking Language Barriers Through Agentic Voice AI

Problem Statement

How can multilingual, real-time voice AI make digital services more accessible across India, enabling people to access information, complete everyday tasks, and interact with essential services naturally in their preferred language?

What Participants Build

A multilingual, voice-first AI agent that understands Indian languages, regional accents, and code-mixed conversations while using tool calling and external integrations to complete meaningful real-world tasks.

Example Solutions

  • Government Services Assistant: Identifies relevant schemes, checks eligibility, explains documentation, and assists with applications.
  • AI Healthcare Assistant: Helps users find doctors, book appointments, access healthcare services, and manage follow-ups.
  • Smart Agricultural Assistant: Provides crop guidance, retrieves mandi prices, checks weather conditions, and connects farmers with relevant services.
  • Multilingual Education Companion: Delivers personalized tutoring, schedules learning sessions, and tracks student progress.
  • Digital Payments & Banking Assistant: Explains financial services, retrieves account information through authorized integrations, and assists with secure transactions.
  • Local Services Assistant: Helps users discover nearby services, schedule appointments, and coordinate service requests.
ps11
Agentic Workflow Opportunity Example: Multilingual Healthcare Assistant Before: Understand the user’s preferred language, location, and healthcare requirements. During: Engage in natural conversations, understand requests, and recommend suitable healthcare providers. Action: Check doctor availability, book appointments, and send confirmations through APIs. After: Deliver appointment reminders, collect feedback, and coordinate follow-up visits. Technology Focus Agora Conversational AI, Agora RTC and/or Agora Chat, Murf AI Falcon TTS (supported languages), multilingual ASR, LLMs, RAG, tool calling, external APIs. Required Demo Flow Participants should demonstrate at least one complete multilingual, agentic service journey: Initiate: User interacts naturally in a supported Indian language through Agora Conversational AI and Murf AI Falcon TTS. Understand: AI interprets user intent, including regional expressions or code-mixed language, and retrieves relevant information. Assist: AI provides contextual guidance and identifies the appropriate service or next action. Execute: AI performs a meaningful action through tool calling or external APIs, such as booking an appointment, submitting a service request, or scheduling an activity. Confirm & Follow Up: AI confirms the outcome in the user’s preferred language and initiates relevant reminders or follow-up actions. Expected Deliverable A working multilingual voice AI assistant demonstrating natural conversations in at least one supported Indian language, contextual understanding, and successful completion of an end-to-end service workflow through external tool/API integrations. Note: Participants are encouraged to explore regional languages, Hinglish and other code-mixed conversations, accessibility-first experiences, and innovative solutions addressing real challenges across urban and rural India. The focus should be on making AI not just multilingual, but genuinely useful and actionable.
Starts18 Nov 202610:00 AM IST
Ends21 Nov 20262:00 PM IST
Interested Members!
Punith VS
Debasis Maharana
Jai Sipani
Aravinthan
Vivan Rajath
ajaypvpv
Hemanth Kumar Reddy

 ... + 199 More