AI WhatsApp Solar Sales Agent
Budget: ₹12,500 – ₹37,500 INR
I run PM Surya Ghar, a residential-solar company that now needs an AI-driven WhatsApp sales agent focused squarely on sales conversion. My in-house team will wire the final output into Zoho CRM, so you can stay laser-focused on WhatsApp automation, AI logic and middleware.
What I need you to build
• Gupshup WhatsApp API integration, fully authenticated under my Business Account.
• A ChatGPT / OpenAI powered conversational layer that replies in a natural mix of Hindi and English—free-text, no rigid button flows.
• Prompt engineering and initial fine-tuning so the agent can:
– greet prospects, collect their city & state, monthly electricity bill and interest level
– qualify or disqualify leads automatically
– hand off high-intent contacts to our human closers through a webhook.
• Scheduled, context-aware follow-ups (e.g., D+1, D+3, D+7) that pick up the thread as a human would rather than sending generic reminders.
• Webhook or lightweight middleware (Node, Python or your preferred stack) that bridges OpenAI, Gupshup and Zoho endpoints.
Acceptance criteria
The agent must hold a multi-turn conversation that ends with a tagged, fully populated lead record (location, bill amount, interest level) pushed to our Zoho insertion endpoint, or an automated marking of “not interested.” Any fallback to a human operator must be logged with the chat transcript.
To help me select quickly, please include:
1. A link or brief on similar AI-WhatsApp projects you’ve shipped.
2. The tech stack you plan to use (language, framework, hosting).
3. A realistic setup timeline from contract sign to first live test.
4. Your total project cost, broken down by milestones if relevant.
Looking forward to seeing how you’ll turn WhatsApp into our most effective solar sales channel.
What I need you to build
• Gupshup WhatsApp API integration, fully authenticated under my Business Account.
• A ChatGPT / OpenAI powered conversational layer that replies in a natural mix of Hindi and English—free-text, no rigid button flows.
• Prompt engineering and initial fine-tuning so the agent can:
– greet prospects, collect their city & state, monthly electricity bill and interest level
– qualify or disqualify leads automatically
– hand off high-intent contacts to our human closers through a webhook.
• Scheduled, context-aware follow-ups (e.g., D+1, D+3, D+7) that pick up the thread as a human would rather than sending generic reminders.
• Webhook or lightweight middleware (Node, Python or your preferred stack) that bridges OpenAI, Gupshup and Zoho endpoints.
Acceptance criteria
The agent must hold a multi-turn conversation that ends with a tagged, fully populated lead record (location, bill amount, interest level) pushed to our Zoho insertion endpoint, or an automated marking of “not interested.” Any fallback to a human operator must be logged with the chat transcript.
To help me select quickly, please include:
1. A link or brief on similar AI-WhatsApp projects you’ve shipped.
2. The tech stack you plan to use (language, framework, hosting).
3. A realistic setup timeline from contract sign to first live test.
4. Your total project cost, broken down by milestones if relevant.
Looking forward to seeing how you’ll turn WhatsApp into our most effective solar sales channel.