Generate personalised WhatsApp cold outreach from scraped leads with OpenAI

Find Your AI Training Job
Chat with DigiNo Toucan and get matched to remote roles in one minute.

B2B sales agencies spend hours researching prospects and writing cold messages one by one. This AI automation handles the full pipeline from raw prospect data to a personalised WhatsApp message, with no manual writing required.

What This Automation Does

Scrapes prospect data from the web using Apify and stores it in a vector database for AI-powered retrieval
Uses a RAG approach to generate highly personalised outreach messages grounded in each prospect's actual business context
Sends the finished messages automatically through WhatsApp, replacing manual outreach with a repeatable, scalable pipeline
Monitors Google Drive for new prospect files, triggering the entire sequence without any manual kick-off

Tools Used

Apify
OpenAI
Google Drive
Supabase
PostgreSQL
n8n

Work faster

Stop Typing and Start Speaking With Typeless

Typeless turns what you say into clean, formatted text in any app on Mac, Windows, Linux, iOS and Android, which it rates at 4x faster than typing. It strips filler words, keeps only your final phrasing when you change your mind mid-sentence, and works in 100+ languages. Emails, AI prompts, client messages and docs, done by voice.

Try Typeless →

Where to Get Hired for This Skill

On Contra, top freelancers across this stack have earned 170 combined verified reviews from real client projects.

Source: Contra freelancer search · refreshed 26 May 2026

Start Earning as a Freelancer on Contra

Contra is a commission-free professional network for independents. Browse live AI automation work and keep what you earn.

Join Contra Free →

How To Build It

Connect Google Drive as the prospect intake point

The workflow watches a designated Google Drive folder so that whenever a client drops in a new prospect list, the entire outreach sequence kicks off automatically without any manual trigger.

Scrape and chunk prospect data with Apify

Apify pulls live web data about each prospect, which is then split into manageable chunks and loaded into a Supabase vector store so the AI can retrieve only the most relevant context when writing each message.

Build a retrieval-augmented prompt for each prospect

For every contact in the list, the workflow queries the vector store to surface the most relevant scraped facts about that prospect, assembling a context-rich prompt that forces the AI to write something specific rather than generic.

Generate and quality-check personalised messages with OpenAI

GPT-4o drafts a cold outreach message for each prospect using the retrieved context, with conversation memory stored in PostgreSQL so the agent can refine tone or length if the client requests adjustments without starting over.

Deliver approved messages through WhatsApp at scale

Once messages pass the AI review step, the workflow dispatches them through the WhatsApp integration in batches, giving the sales agency a fully hands-off outbound channel that runs without any human sending effort.

Pitfalls

Apify scrapers break when target websites change their HTML structure, so builds delivered to clients need a monitoring alert that flags failed scrape runs before the client notices gaps in their outreach.
WhatsApp Business API accounts can be flagged or rate-limited if message volume spikes too fast; batching sends and warming up new accounts gradually is essential to keep the pipeline running.
Supabase vector store embeddings become stale if prospect data is re-scraped but old vectors are not deleted first, causing the AI to retrieve outdated context and produce inaccurate personalisation.

FAQ

Can I build this without coding?

Yes. The workflow uses a visual low-code environment throughout, and the most technical parts, such as connecting Supabase and configuring Apify actors, are done through form fields and API keys rather than written code. Comfort with APIs and reading documentation will help you move faster.

How long does it take?

Expect 8 to 15 hours for a first build, including testing with a real prospect list. The Apify and WhatsApp API setup tends to take the longest because both require account approval and configuration outside the workflow itself.

What can I charge?

Pricing is your call based on your positioning and the client's outreach volume. Most freelancers offering done-for-you outbound pipeline builds price on project value, such as the number of prospects processed or messages delivered per month, rather than by the hour.

Which tool is required vs optional?

OpenAI and Apify are core and cannot be swapped out without rebuilding significant logic. Google Drive is optional as the intake trigger and can be replaced with a form, webhook, or Google Sheets upload. Supabase can in principle be replaced with another vector store, but it is the most straightforward option given existing n8n support.

This is original DigiNo analysis. The underlying automation pattern is a community workflow template – view the original on n8n.