AI-Powered Instagram Real Estate Lead Generator
Budget / Salary$250–750
TypeFreelance project
LocationRemote
Posted1 hour ago
# AI Real Estate Instagram Lead Generation System — Build From Scratch
I am looking for an experienced AI/automation developer to design and build a complete Instagram DM real-estate lead generation system **from scratch**.
The business focuses initially on **Abu Dhabi real estate**, with Dubai expansion planned later.
The goal is not to build a basic chatbot or a fixed decision-tree automation.
The goal is to build an intelligent real-estate sales assistant that can answer project questions using verified data, understand natural buyer conversations, recommend relevant projects, qualify buyers, identify high-intent leads, and send structured leads to a CRM for human follow-up.
## Scope
The freelancer will be responsible for proposing, designing and implementing the complete technical architecture.
I am open to ManyChat, Botpress, custom development, APIs, databases, CRM platforms, or a combination.
I do not want the project designed around a particular tool unless there is a good technical/business reason for using it.
The finished system must be manageable by a non-developer after handover.
## 1. Instagram AI Assistant
Customers will interact with the system through Instagram DMs.
The assistant must understand natural messages such as:
* “I have 3M”
* “What can I buy on Yas?”
* “Any new projects?”
* “Show me villas”
* “Anything from Aldar?”
* “I have 500k cash now”
* “I need a payment plan”
* “What can I get in Hudayriyat?”
* “I want a 3 bedroom”
* “What is available?”
* “I want to reserve”
The conversation should feel like speaking with a knowledgeable UAE property adviser rather than completing a chatbot questionnaire.
The system must maintain conversation context and avoid repeatedly asking for information the customer has already provided.
## 2. Real Estate Knowledge Database
The freelancer must build a structured and easily editable property/project database.
It should support information such as:
* Developer
* Project
* Emirate
* Area/location
* Property type
* Bedroom options
* Starting prices
* Sizes
* Payment plans
* Required initial payment
* Handover
* Project status
* Availability where available
* Project description/features
* Source
* Last verified/updated date
I must be able to add projects, change prices, update payment plans and modify information without changing AI prompts or rebuilding automations.
### No Hallucinated Property Data
This is a critical requirement.
Prices, availability, payment plans, handover dates and other commercial/project facts must come from approved data.
The AI must never invent, estimate or assume missing property information.
If information cannot be verified, the assistant should clearly say so while continuing the conversation and helping/qualifying the buyer where possible.
Missing data alone should not automatically trigger a human handoff.
## 3. Intelligent Property Matching
The assistant should use known buyer requirements to find relevant properties/projects.
Examples include matching by:
* Budget
* Available initial cash
* Location
* Developer
* Property type
* Bedrooms
* Investment/end use
* Payment preference
* Purchase timeframe
Recommendations must respect the actual database.
If a project does not have studios, for example, the assistant must not recommend a studio there.
## 4. Buyer Qualification
Qualification should happen naturally throughout the conversation rather than through a long form.
The system should progressively identify useful information including:
* Budget
* Available cash/down payment
* Preferred location
* Project/developer interest
* Property type
* Bedrooms
* Investment or end use
* Cash / mortgage / developer payment plan
* Purchase timeframe
* Name
* Phone/WhatsApp when appropriate
The assistant should provide useful information before aggressively requesting personal contact information.
It should never repeatedly ask questions already answered during the conversation.
## 5. Lead Intent Detection
The system should recognize strong buying signals.
Examples include:
“Reserve”
“Book”
“EOI”
“Availability”
“Payment plan”
“Ready to buy”
“WhatsApp me”
“Speak to an agent”
High-intent prospects should be clearly flagged for rapid human follow-up.
## 6. CRM
The freelancer should recommend and implement an appropriate CRM integration.
Qualified leads should be stored as structured records containing relevant information such as:
* Contact information
* Instagram source/identity where available
* Budget
* Available cash
* Location
* Developer/project interest
* Property requirements
* Financing preference
* Purchase timeframe
* Intent signals
* Conversation summary
* Lead status
* Follow-up status
The system should avoid unnecessary duplicate contacts and should update existing leads where appropriate.
## 7. Human Handoff
Human involvement should occur when useful, particularly for strong purchase intent, reservation requests, confirmed availability requests or when the customer explicitly asks for an agent.
The AI should not unnecessarily transfer conversations simply because some information is unavailable.
## 8. Data Management
This is extremely important.
After handover, a non-technical user must be able to:
* Add a developer
* Add a project
* Change a price
* Update a payment plan
* Change availability/status
* Correct project information
* Disable outdated information
The freelancer should propose the simplest reliable method for managing this data.
## 9. Future Expansion
The initial architecture should allow future expansion to:
* WhatsApp
* Website chat
* Dubai properties
* Automated follow-up
* New project alerts
* Property/news content
* Instagram content automation
* Lead re-engagement
* Additional sales agents
These do not all need to be included in Phase 1, but the architecture should not make future expansion unnecessarily difficult.
## Deliverables
The project should include:
1. Technical architecture and setup
2. Instagram DM integration
3. AI conversational system
4. Property/project database
5. Verified-data retrieval system
6. Property matching logic
7. Buyer qualification
8. Lead-intent detection
9. CRM implementation/integration
10. Human handoff workflow
11. Error and missing-data handling
12. Testing
13. Deployment
14. Documentation
15. Full handover
All custom code, prompts, automations, databases, API configurations and documentation created for the project must be handed over at completion.
The final system and relevant accounts/infrastructure must be under my control.
## Acceptance Testing
The completed system will be tested using realistic buyer conversations, including:
“I have AED 3M.”
“What can I buy in Hudayriyat?”
“Show me new projects.”
“Anything from Aldar?”
“I have 500k now and need a payment plan.”
“I want a studio.”
“I want a villa.”
“What is the payment plan?”
“What is available?”
“I want to reserve.”
“WhatsApp me.”
The system will also be tested with missing data, unknown projects, returning customers and customers who do not want to provide contact details.
## What I Want From the Freelancer
Please do **not** send a generic chatbot proposal.
In your bid, explain:
1. The architecture you recommend.
2. The technology/platforms you would use and why.
3. How Instagram messaging will connect to the AI.
4. How property information will be stored and maintained.
5. How you will prevent the AI from inventing prices, payment plans or availability.
6. How property matching will work.
7. How leads will be qualified and stored.
8. Which CRM you recommend and why.
9. Examples of similar systems you have built.
10. Estimated development time.
11. Fixed project cost.
12. Any recurring monthly/API costs expected after launch.
Start your proposal with **“UAE REAL ESTATE AI”** so I know you have read the complete specification.
## Success Criteria
The system should turn:
**Instagram DM → intelligent conversation → verified property information → relevant recommendations → buyer qualification → high-intent detection → structured CRM lead → human sales follow-up**
The primary business objective is **qualified real-estate leads and sales opportunities**, not chatbot engagement metrics.
I am looking for an experienced AI/automation developer to design and build a complete Instagram DM real-estate lead generation system **from scratch**.
The business focuses initially on **Abu Dhabi real estate**, with Dubai expansion planned later.
The goal is not to build a basic chatbot or a fixed decision-tree automation.
The goal is to build an intelligent real-estate sales assistant that can answer project questions using verified data, understand natural buyer conversations, recommend relevant projects, qualify buyers, identify high-intent leads, and send structured leads to a CRM for human follow-up.
## Scope
The freelancer will be responsible for proposing, designing and implementing the complete technical architecture.
I am open to ManyChat, Botpress, custom development, APIs, databases, CRM platforms, or a combination.
I do not want the project designed around a particular tool unless there is a good technical/business reason for using it.
The finished system must be manageable by a non-developer after handover.
## 1. Instagram AI Assistant
Customers will interact with the system through Instagram DMs.
The assistant must understand natural messages such as:
* “I have 3M”
* “What can I buy on Yas?”
* “Any new projects?”
* “Show me villas”
* “Anything from Aldar?”
* “I have 500k cash now”
* “I need a payment plan”
* “What can I get in Hudayriyat?”
* “I want a 3 bedroom”
* “What is available?”
* “I want to reserve”
The conversation should feel like speaking with a knowledgeable UAE property adviser rather than completing a chatbot questionnaire.
The system must maintain conversation context and avoid repeatedly asking for information the customer has already provided.
## 2. Real Estate Knowledge Database
The freelancer must build a structured and easily editable property/project database.
It should support information such as:
* Developer
* Project
* Emirate
* Area/location
* Property type
* Bedroom options
* Starting prices
* Sizes
* Payment plans
* Required initial payment
* Handover
* Project status
* Availability where available
* Project description/features
* Source
* Last verified/updated date
I must be able to add projects, change prices, update payment plans and modify information without changing AI prompts or rebuilding automations.
### No Hallucinated Property Data
This is a critical requirement.
Prices, availability, payment plans, handover dates and other commercial/project facts must come from approved data.
The AI must never invent, estimate or assume missing property information.
If information cannot be verified, the assistant should clearly say so while continuing the conversation and helping/qualifying the buyer where possible.
Missing data alone should not automatically trigger a human handoff.
## 3. Intelligent Property Matching
The assistant should use known buyer requirements to find relevant properties/projects.
Examples include matching by:
* Budget
* Available initial cash
* Location
* Developer
* Property type
* Bedrooms
* Investment/end use
* Payment preference
* Purchase timeframe
Recommendations must respect the actual database.
If a project does not have studios, for example, the assistant must not recommend a studio there.
## 4. Buyer Qualification
Qualification should happen naturally throughout the conversation rather than through a long form.
The system should progressively identify useful information including:
* Budget
* Available cash/down payment
* Preferred location
* Project/developer interest
* Property type
* Bedrooms
* Investment or end use
* Cash / mortgage / developer payment plan
* Purchase timeframe
* Name
* Phone/WhatsApp when appropriate
The assistant should provide useful information before aggressively requesting personal contact information.
It should never repeatedly ask questions already answered during the conversation.
## 5. Lead Intent Detection
The system should recognize strong buying signals.
Examples include:
“Reserve”
“Book”
“EOI”
“Availability”
“Payment plan”
“Ready to buy”
“WhatsApp me”
“Speak to an agent”
High-intent prospects should be clearly flagged for rapid human follow-up.
## 6. CRM
The freelancer should recommend and implement an appropriate CRM integration.
Qualified leads should be stored as structured records containing relevant information such as:
* Contact information
* Instagram source/identity where available
* Budget
* Available cash
* Location
* Developer/project interest
* Property requirements
* Financing preference
* Purchase timeframe
* Intent signals
* Conversation summary
* Lead status
* Follow-up status
The system should avoid unnecessary duplicate contacts and should update existing leads where appropriate.
## 7. Human Handoff
Human involvement should occur when useful, particularly for strong purchase intent, reservation requests, confirmed availability requests or when the customer explicitly asks for an agent.
The AI should not unnecessarily transfer conversations simply because some information is unavailable.
## 8. Data Management
This is extremely important.
After handover, a non-technical user must be able to:
* Add a developer
* Add a project
* Change a price
* Update a payment plan
* Change availability/status
* Correct project information
* Disable outdated information
The freelancer should propose the simplest reliable method for managing this data.
## 9. Future Expansion
The initial architecture should allow future expansion to:
* Website chat
* Dubai properties
* Automated follow-up
* New project alerts
* Property/news content
* Instagram content automation
* Lead re-engagement
* Additional sales agents
These do not all need to be included in Phase 1, but the architecture should not make future expansion unnecessarily difficult.
## Deliverables
The project should include:
1. Technical architecture and setup
2. Instagram DM integration
3. AI conversational system
4. Property/project database
5. Verified-data retrieval system
6. Property matching logic
7. Buyer qualification
8. Lead-intent detection
9. CRM implementation/integration
10. Human handoff workflow
11. Error and missing-data handling
12. Testing
13. Deployment
14. Documentation
15. Full handover
All custom code, prompts, automations, databases, API configurations and documentation created for the project must be handed over at completion.
The final system and relevant accounts/infrastructure must be under my control.
## Acceptance Testing
The completed system will be tested using realistic buyer conversations, including:
“I have AED 3M.”
“What can I buy in Hudayriyat?”
“Show me new projects.”
“Anything from Aldar?”
“I have 500k now and need a payment plan.”
“I want a studio.”
“I want a villa.”
“What is the payment plan?”
“What is available?”
“I want to reserve.”
“WhatsApp me.”
The system will also be tested with missing data, unknown projects, returning customers and customers who do not want to provide contact details.
## What I Want From the Freelancer
Please do **not** send a generic chatbot proposal.
In your bid, explain:
1. The architecture you recommend.
2. The technology/platforms you would use and why.
3. How Instagram messaging will connect to the AI.
4. How property information will be stored and maintained.
5. How you will prevent the AI from inventing prices, payment plans or availability.
6. How property matching will work.
7. How leads will be qualified and stored.
8. Which CRM you recommend and why.
9. Examples of similar systems you have built.
10. Estimated development time.
11. Fixed project cost.
12. Any recurring monthly/API costs expected after launch.
Start your proposal with **“UAE REAL ESTATE AI”** so I know you have read the complete specification.
## Success Criteria
The system should turn:
**Instagram DM → intelligent conversation → verified property information → relevant recommendations → buyer qualification → high-intent detection → structured CRM lead → human sales follow-up**
The primary business objective is **qualified real-estate leads and sales opportunities**, not chatbot engagement metrics.
Apply on Freelancer →
Project sourced from Freelancer.com. Applications happen directly on the original platform — we never collect your data.