AI agent or process automation, built into your systems by Pablo ArcaAI agent or process automation, built into your systems by Pablo Arca
AI agent or process automation, built into your systemsPablo Arca
Cover image for AI agent or process automation, built into your systems
You get an AI agent or an automated process that works inside your own tools (WhatsApp, your database, your spreadsheet). Not a demo.
I start with a two-week paid pilot, and its price comes off the full project. Before delivery I collect real cases with the right answer written next to each one, and every version has to pass them. In production I keep watching each run in Arize Phoenix, with cost and latency.
The model reads and proposes. My code checks its output with Pydantic and decides what reaches your systems. Business rules live in the code, not in the prompt. Repository, Docker and deploys belong to you, and model and hosting costs go on your own accounts.
Some of what I have built: a WhatsApp agent for a real estate agency that gave the team three hours a day back. A booking assistant for clinics that handles health data under GDPR. Half a dozen specialised agents behind one single entry point for a large industrial company (Cosentino), with the routing in code.
After launch I can run and monitor it for a monthly fee, with no lock-in.
FAQs

Starting at$1,750
Duration2 weeks
Tags
FastAPI
OpenAI
Python
WhatsApp
AI Agent Developer
AI Automation
AI Chatbot Developer
Business Workflow Automation
Service provided by
Pablo Arca Málaga, Spain
AI agent or process automation, built into your systemsPablo Arca
Starting at$1,750
Duration2 weeks
Tags
FastAPI
OpenAI
Python
WhatsApp
AI Agent Developer
AI Automation
AI Chatbot Developer
Business Workflow Automation
Cover image for AI agent or process automation, built into your systems
You get an AI agent or an automated process that works inside your own tools (WhatsApp, your database, your spreadsheet). Not a demo.
I start with a two-week paid pilot, and its price comes off the full project. Before delivery I collect real cases with the right answer written next to each one, and every version has to pass them. In production I keep watching each run in Arize Phoenix, with cost and latency.
The model reads and proposes. My code checks its output with Pydantic and decides what reaches your systems. Business rules live in the code, not in the prompt. Repository, Docker and deploys belong to you, and model and hosting costs go on your own accounts.
Some of what I have built: a WhatsApp agent for a real estate agency that gave the team three hours a day back. A booking assistant for clinics that handles health data under GDPR. Half a dozen specialised agents behind one single entry point for a large industrial company (Cosentino), with the routing in code.
After launch I can run and monitor it for a monthly fee, with no lock-in.
FAQs

$1,750