Jeffrey Quemuel
Hi, I’m Jeffrey. I design AI-powered automation systems that replace repetitive work and scale your business faster.
From n8n workflows to RAG-based AI assistants, email marketing engines, and Facebook chatbots .I build systems that run your operations on autopilot.
If it can be automated, I’ll build it. If it can be optimized, I’ll improve it.
CASE STUDY :
Email Automation Without Domain Access (Real Client Case)
When email deliverability broke in 7 days — I rebuilt the system without touching the domain.
🧩 The Problem
Client wanted automated outreach: 300–400 emails/day
Within 1 week → emails went straight to spam
No proper SPF, DKIM, or DMARC setup
Blocked access to domain & website setting
👉 Result: Broken deliverability + no control over core fixes
🚧 The Constraint
Most developers stop here.
No domain access =
No reputation repair
No authentication fixes
No traditional solution
The Solution
Instead of fixing the domain, I rebuilt the sending system from scratch using Python
✔ Custom email automation engine
✔ Controlled sending intervals
✔ Human-like behavior simulation
✔ Smart batching to avoid spam triggers
✔ Fully independent from website infrastructure
How It Works
Sends emails gradually, not in bulk spikes
Randomized delays → avoids bot detection
Daily limits maintained without overload
Clean execution logic using Python
The Result
✔ Stabilized email sending
✔ Continued 300–400 emails/day safely
✔ Reduced spam issues without domain fixes
✔ Delivered a working system under restrictions
AI Chatbot Upgrade with RAG (MedSpa)
From Basic Chatbot to Smart AI Assistant (RAG Implementation)
🧩 The Problem
A MedSpa client was already using an AI chatbot connected to Google Docs.
On paper, it worked. In reality:
AI missed important details from documents
Responses were inconsistent and incomplete
Information retrieval was unreliable
Client experience felt “hit or miss”
👉 The chatbot wasn’t trustworthy for real customer interactions
🚧 The Limitation
The system relied only on:
Google Docs as the knowledge base
A basic AI agent without structured retrieval
👉 No intelligent search layer
👉 No proper context handling
Result: AI guessing instead of knowing
💡 The Solution
I upgraded the system by implementing a Retrieval-Augmented Generation (RAG) pipeline
✔ Integrated Supabase as a vector database
✔ Connected Google Drive as the document source
✔ Enabled semantic search for accurate retrieval
✔ Structured data flow for better AI context
👉 Now the AI doesn’t guess — it retrieves and responds
⚙️ How It Works
Google Drive → Supabase (Vector DB) → AI Agent → Accurate Response
Documents are stored and synced from Google Drive
Data is processed into embeddings
Supabase stores and retrieves relevant context
AI generates answers based on precise data
📈 The Result
✔ More accurate and consistent answers
✔ Reduced hallucinations
✔ Faster response quality improvement
✔ Better customer experience for MedSpa clients
👉 The chatbot became a reliable assistant, not a risky tool
TOOLS :
N8N
GOOGLE APPS
AIRTABLE
RAG
JAVASCRIPT
PYTHON