Built a Full Retrieval-Augmented Generation (RAG) Workflow on Make.com (http://Make.com)
I Just wrapped up a proof-of-concept for an internal Support Agent Assistant, an AI that answers complex product questions using only the company’s own documentation.
Stack overview:
1. Make.com (http://Make.com) for the entire RAG pipeline
2. Render (Python Flask) as a serverless ingestion endpoint
3. Weaviate as the vector database
4. AI by Make for grounded text generation
Zero manual prompts, one smart workflow. ⚡
#AI #Automation #RAG #Makecom #Weaviate #NoCode #Python #Render #LLM
5
6
367
AI Inbox Assistant: (Make + GPT)
I built an AI inbox assistant that reads incoming emails, classifies them, suggests a one-line reply, and follows up on anything older than 48 hours.
Why it works: faster responses, fewer dropped threads, VIPs prioritized, and clear weekly visibility, engineered with strict JSON outputs and ISO timestamps for easy upkeep.
Want this for your inbox? I’ll tailor it to your pipeline, VIP list, and KPIs.
#AI #Automation #NoCode #GPT #WorkflowAutomation #CRM #SaaS
0
292
🚀 Built an AI Email Filtering System that Saves 8+ Hours a Day
I’ve been building AI agents that cut out manual work, and this one made a real difference.
A client’s inbox was packed with spam, pitches, invoices, and real inquiries all mixed together. The team spent hours sorting every morning.
So I built an AI-powered email filter with Make, OpenAI (GPT-5), and Airtable.
Each incoming email is instantly tagged as:
✅ Qualified Inquiry
✅ Spam / Irrelevant
✅ Invoice
How it works:
• Make connects Gmail, OpenAI, and Airtable in real time.
• GPT-5 analyzes each message to detect intent.
• Airtable logs everything for visibility.
Results:
💡 95% less spam in the main inbox
💡 8+ hours saved daily
💡 Instant focus on real clients
#Automation #casestudy #MakeCom #EmailAutomation #AIAgent #Productivity #ContraCommunity #NoCode #OpenAI