AI-Powered Offer Automation MVP (Make.com or n8n + OpenAI)

AI-Powered Offer Automation MVP (Make.com or n8n + OpenAI) - pref. with make.com as i have already an existing account there

Description

We are looking for an experienced Make.com or n8n automation expert to build an AI-based offer generation workflow as part of a professional training project.
This will be an MVP (Minimum Viable Product) to demonstrate the concept and can be reused for multiple clients later.

The workflow should capture and process customer inquiries from several channels (WhatsApp, Telegram, e-mail, and iPhone voice messages), analyze them with OpenAI, check a rule-based FAQ, retrieve data from structured sources (Excel or Airtable), and generate a draft offer or e-mail reply.

Scope of Work

1. Input Handling

Accept text and voice inputs (voice from iPhone or WhatsApp).

Use OpenAI Whisper for automatic transcription.

Handle messages via WhatsApp, Telegram, or e-mail (IMAP/Gmail/Outlook).

2. AI-Driven Analysis

Use OpenAI GPT-4 or GPT-4o for message understanding and intent detection.

Match incoming requests against an FAQ or ruleset stored in Airtable (or Excel if Airtable costs are too high).

Determine if the response should be an FAQ answer, a generated offer, or a manual escalation.

3. Data Sources (per client)
Each client has their own dataset:

/Customer_A/
- price_list.xlsx
- customer_list.xlsx
- faq.xlsx


or alternatively, an Airtable base containing:

Customer data

Product and pricing data

FAQ / rule logic

4. Processing and Output

Retrieve customer and product information.

Combine and generate a structured offer or e-mail in German.

If a template is available, create a Word or PDF offer; otherwise, generate a formatted e-mail text.

Optional manual quality review before sending.

Send e-mail automatically or after approval.

5. Storage and Logging

Save all inquiries, responses, and offers.

Allow export of logs to Airtable, Google Sheets, or a shared directory.

FAQ / Rule Framework (Airtable)

Each rule entry should include:

Category

Trigger Keywords

Instruction for AI

Default Response

Escalation Required (True/False)

If Airtable is not used, the same logic can be implemented in Excel files stored in a client-specific folder.

Technical Requirements

Make.com or n8n

OpenAI GPT-4 / GPT-4o for natural-language analysis

OpenAI Whisper for transcription

Airtable or cloud file storage (Google Drive, Dropbox, or OneDrive)

Output: Word (.docx) or PDF

E-mail integration: Gmail, Outlook, or SMTP

Modular and reusable design for multiple clients

Deliverables

Reusable Make.com or n8n workflow (importable)

Example Airtable base or folder structure with sample data

Documentation for setup and client configuration

Demonstration with 2–3 test inputs (text, e-mail, and voice)

Scope: MVP phase, no full system integrations yet

Data can be stored in Airtable or as Excel files in shared folders

Future extensions for ERP/CRM integrations possible

Required Skills

Make.com or n8n

OpenAI API integration

Whisper speech-to-text

Airtable or data automation experience

Workflow design and testing