01/ case study
AI Invoice Processor
Turns PDF and image invoices into structured Excel across 200+ templates.
PythonOpenAI APIOCRPandas
Available for new projects · Q3 2026
Production-ready AI.
Not experiments.
I'm Antonio — a freelance AI engineer building reliable LLM applications, automations, and full-stack products for SMEs and agencies. From first call to shipped software that earns its keep.
Shipped
production AI apps
Stack
Python · Next.js · LLMs
Background
Ericsson · Photomath
Approach
reliability-first
Services
What I actually do
Four ways I help teams get real value from AI — without the buzzword detour.
LLM Application Development
AI assistants, chatbots, and document tools — taken from rough idea to production software your team actually uses.
AI Process Automation
Cut manual work in invoicing, lead generation, and customer queries with LLM pipelines built around your existing tools.
API Integration & Consulting
Connect your current systems to AI, advise on adoption, and untangle proof-of-concepts that never made it to production.
Full-Stack Delivery
End-to-end web apps with AI features — Next.js, Python, deployed and maintained, not handed off as a half-built repo.
Selected work
Things I've shipped
Real projects, real outcomes. Tech is a tool — the result is the point.
02/ case study
AI Lead Generator
Scraping + LLM pipeline that cleans and enriches business data — cut manual prospecting by ~80%.
PythonOpenAI APIBeautifulSoup
03/ case study
AI Receptionist
Voice assistant with intent detection and context-aware routing via the Realtime API.
PythonRealtime APIFastAPI
04/ case study
StarApp
Full-stack natal-chart web app with LLM-generated interpretations, shipped to production.
Next.jsPythonOpenAI APIVercel
05/ case study
Tidio Chatbot Integration
Customized chatbot integrated into a client site with tailored conversation flows.
TidioJavaScriptIntegration
10+
solo AI projects designed and shipped
4+
years building production software
100K+
records processed in live data pipelines
200+
document formats automated end-to-end
About
Reliability-first engineering, applied to AI.
My background isn't another "AI startup founder." It's hands-on data and AI work — building pipelines that handle millions of records, structuring the training data behind real products, and building LLM applications end to end. The focus is always the same: not a flashy demo, but software reliable enough to run a business on.
I work with one client at a time, end to end. You talk to the engineer building it — not an account manager.
- Big Data Analyst at Ericsson Nikola Tesla — large-scale telecom data systems where uptime isn't optional.
- Data annotation at Photomath (acquired by Google) — structuring 100K+ training samples behind a math app used by millions.
- Lead Developer on a real-time control system for a special-purpose vehicle — software that has to work, every time.
Process
How we work together
A short, predictable path from first call to shipped product.
01
Discovery
Short call to understand the problem, the constraints, and what success actually looks like.
02
Build
Tight iterations with working software you can try, not slide decks or mockups.
03
Ship
Deployed to production with monitoring, docs, and a clear handover.
04
Support
Ongoing improvements and reliability work — you keep a direct line to the engineer who built it.
Tech stack
PythonFastAPIOpenAIAnthropicNext.jsSQLVercelLinux
Let's talk
Book a free consultation.
Tell me what you're trying to build or fix. I'll reply within one business day with honest feedback — and whether I'm the right person for it.