CV
AI Engineer Automation Developer

Mateo Costantini

Developing the future with Artificial Intelligence

5th year Artificial Intelligence Engineering student at San Andrés University. Passionate about the intersection of technology, business, and innovation, with a strong focus on developing scalable solutions using advanced AI, LLMs, and Computer Vision to optimize processes and solve real-world problems.

Contact Me Download CV
Mateo Costantini

Autonomous Agents

Integrating modern LLMs

Smart Workflows

Scaling business operations

Scroll
Profile

About Me

I build and deploy intelligent automation systems and machine learning workflows. My academic path merges rigorous mathematics, software engineering, and artificial intelligence, enabling me to translate complex operational needs into scalable, high-impact products.

AI & Deep Learning

Designing cognitive architectures and training neural models for computer vision and text analysis.

  • PyTorch & Deep Learning
  • LLMs & RAG Pipelines
  • Computer Vision & NLP

Process Automation

Building secure, high-volume automated scripts to eliminate manual operations and structure data flow.

  • Python Scripting & DB Ingestion
  • Automated Reporting & ETL
  • Docker & CI/CD Pipelines

Academic & Leadership

Striving for technical excellence, coordinating teams, and driving social impact projects.

  • 1st Place AI Challenge
  • Cambridge Advanced C1
  • Youth Leadership & Coaching
5th Year AI Engineering Student
2+ Years Professional Experience
7.84 UdeSA Academic GPA
Portfolio

Featured Projects

Fullstack & Microservices

Udesamigos

A university social platform built on a microservices architecture. Implemented a distributed backend (Node.js/Express, Python/FastAPI), centralized authentication, push notification integration using Firebase Cloud Messaging (FCM), and containerization with Docker.

React Native Node.js Python Firebase Docker
Private Microservices
Deep Learning / PyTorch

Co2L Continual Learning

Implementation and evaluation of pipelines for Class-Incremental Learning (Class-IL). Benchmarked Naive, EWC, LwF, and Co2L algorithms utilizing supervised contrastive representation loss (SupCon) and Nearest Class Mean (NCM) classifiers.

PyTorch Python SupCon NCM Continual Learning
Academic Advanced AI
Applied AI / LLMs

Smart OCR Pipeline

Intelligent document and invoice extraction pipeline. Features custom text-cleaning heuristics using PyMuPDF (Fitz) and optimized prompt engineering with LLM APIs for structured data extraction and field-level confidence scoring.

Python LLM API PyMuPDF OCR Data Extraction
Corporate Automation
Expertise

Technical Skills

Deep Learning
Advanced
Computer Vision
Advanced
NLP & LLMs (RAG/Agents)
Advanced
PyTorch
Expert
Python
Expert
C / C++ / Java
Competent
SQL & DB Systems
Advanced
Automation (Workflows)
Expert
AWS / Azure
Intermediate-Advanced
CI/CD & GitHub Actions
Advanced
Docker & Containerization
Advanced
1st Place AI Challenge
San Andrés University
Cambridge Advanced C1
Official Certificate
Leadership & Coordination
Youth Group (2022-2025)
Timeline

Experience & Education

2026 - Present

AI Engineer

Luno

Leading the development and integration of AI-powered solutions into client production workflows.

  • Designed and deployed autonomous agents and advanced RAG workflows utilizing state-of-the-art LLMs to improve data-driven decision-making.
  • Analyzed client business workflows to translate complex commercial requirements into clear technical specifications and functional AI prototypes.
  • Optimized and automated operational processes, achieving a significant reduction in operational costs and manual task times.
2024 - 2026

Automation Developer

Intec Global Solutions

Specialized in creating and maintaining automated reporting workflows and data integrations.

  • Designed and deployed automated workflows that structure and extract high-volume operational reports, improving data accessibility and consistency.
  • Developed Python scripts for scheduled data cleaning and ingestion into internal databases, reducing manual entry errors to zero.
  • Collaborated with cross-functional teams to design business automation architectures at scale.
Interactive

Mateo AI Terminal

Interact with a simulated version of Mateo to learn more about his technical and academic profile. Type a command or click the quick action buttons below.

mateo-ai-agent --bash
online
[SYSTEM] Mateo AI Agent v1.0.0 initialized.
Type help to see available commands.
guest@mateo-ai:~#
guest@mateo-ai:~#
Connect

Get In Touch

Interested in working together?

I am always open to discussing AI engineering opportunities, smart automation workflows, technical consulting, or development roles.

Email Address mateo.costantini.t@gmail.com
LinkedIn linkedin.com/in/mateocostantini
Location Buenos Aires, Argentina