AI-Powered Data Analysis Agent Development -- 2

What I'm Looking For:

We're looking for an LLM-based conversational AI expert who has proven experience building complex, context-aware AI agents – especially for WhatsApp-style group communication.

This is not a simple integration job. The main challenge is designing and training an agent that can replace real human operators who currently provide 24/7 service.
Project Overview:

We run a real estate company in Israel called DiffeRent.
Right now, two human employees work 24/7 to manage communication with hundreds of brokers in over 100 WhatsApp groups.

In these groups, brokers send:

Partial or messy data about potential tenants (sometimes as photos)
Requests for credit checks
Follow-up questions
Updates on deals
The process currently involves:

Extracting tenant info from unstructured WhatsApp messages (text or image)
Entering the info into our Zoho CRM
Waiting for a financial background check result
Sending a professional response back to the group
Following up regularly to get full landlord and apartment details
Converting the case into a clean lead for the sales team to close the deal
Goal:

We want to build a smart AI agent that fully replaces this manual process, manages dozens of conversations in parallel, and handles everything automatically.

The agent must:

Understand WhatsApp group context (multi-user chat)
Extract structured data from messages/images
Respond in a friendly and professional tone
Trigger follow-ups if brokers don’t respond
Push complete cases to Zoho CRM
What You MUST Have:

Proven experience building complex AI agents, not just scripts or bots
Experience with prompt engineering for LLMs (GPT, Claude, etc.)
Ability to handle multi-turn memory, context tracking, and user intent
Bonus: WhatsApp Business / Twilio / Zoho experience
Must Provide:
Concrete examples of real-world agents you've built – not just theory
Links, videos, GitHub repos, or deployed apps showing your AI agent in action
The more advanced the example – the better