DealPilot AI
Link to open source: https://github.com/BlueeWhale/Ecosphere-hackathon
Link to Live Project: https://dealpilot-frontend.onrender.com
DealPilot AI
DealPilot is an AI-powered sales and negotiation platform that combines real-time conversational voice AI with sales intelligence.
It enables an AI sales agent to conduct natural voice conversations with customers, understand their requirements, qualify leads, answer product questions, handle objections, compare competitors, and dynamically adapt its sales strategy.
The platform provides a complete sales workflow through a centralized dashboard, including customer accounts, leads, conversations, active and qualified deals, pipeline value, follow-ups, and Deal State with customer buying intent.
Authentication & Role-Based Access
DealPilot provides separate login access for Admin and Company Users, with role-based permissions and dedicated dashboards.
Admin Login: admin@dealpilot.ai | Password: Admin@2026Secure!
The Admin can manage companies/users, monitor overall sales analytics, view deals, leads and conversations, and manage system-level configurations.
Company User Login: company@dealpilot.ai | Password: Company@2026Secure!
Company Users can manage customers and leads, view conversations, track active and qualified deals, monitor pipeline value, manage follow-ups, and view Deal State and customer buying intent.
Smart Human Handoff
One of DealPilot's key features is Smart Human Handoff. When a customer requires human assistance, the AI can transfer the conversation while preserving the complete context, including conversation history, customer requirements, objections, pricing discussion, competitor information, AI summary, recommendation, Deal State, buying intent, and reason for handoff.
This allows the human representative to continue the conversation without asking the customer to repeat their information.
AI Price & Competitor Comparator
DealPilot can compare the company's offering with competitors when a customer mentions an alternative such as Salesforce. The AI compares pricing, plans, features, customer requirements, and overall value, allowing the salesperson to make a data-driven response instead of immediately offering a discount.
Dynamic Deal State
DealPilot continuously maintains the Deal State throughout the conversation, tracking customer requirements, budget, buying intent, objections, competitors, pricing discussions, deal stage, and conversation history. When requirements change, the Deal State is updated and the AI adapts its sales strategy accordingly.
Technology Stack
-
Agora Conversational AI + Agora RTC — real-time voice communication between customers and the AI agent
-
Google Gemini 2.5 Flash — conversational intelligence, requirement understanding, objection handling, competitor analysis, and adaptive responses
-
MongoDB — persistent storage for users, companies, customers, leads, conversations, deals, and sales data
-
Socket.IO / WebSockets — real-time communication and live updates
-
React + Vite — frontend interface and dashboard
-
Node.js + Express — backend and API layer
-
FastAPI — AI service layer
Core USP
DealPilot doesn't just generate responses. It understands the deal, remembers customer context, compares competitive alternatives, adapts its negotiation strategy, and knows when to bring a human into the conversation.
This build was uploaded as a hackathon project




