class PabloZizzutti:
    def __init__(self):
        self.role = "Senior Software Engineer"
        self.experience = "15+ years"
        self.specialization = [
            "LLM-powered applications",
            "Real-time inference pipelines", 
            "High-performance APIs",
            "Cloud-scale data systems"
        ]
        self.current_focus = ["AI/ML Systems", "ETL Pipelines", "Backend Architecture"]
        self.teaching = "Graduate Mentor @ UBA"
        
    def expertise(self):
        return {
            "ai_ml": [
                "LLM Integration", "RAG", "Prompt Engineering",
                "Vector Databases", "NLP", "Speech Recognition (Whisper)",
                "PyTorch", "TensorFlow", "CUDA", "Generative AI"
            ],
            "backend": ["FastAPI", "Django", "Flask", "GraphQL", "Node.js"],
            "cloud_aws": ["Lambda", "S3", "Glue", "Redshift", "DynamoDB"],
            "cloud_gcp": ["Cloud Run", "BigQuery", "Cloud SQL"],
            "databases": ["PostgreSQL", "MongoDB", "Redis", "DynamoDB"],
            "devops": ["Docker", "Kubernetes", "CI/CD", "DataDog", "Sentry"],
            "embedded": ["C/C++", "STM32", "RTOS", "Raspberry Pi", "IoT"]
        }
    
    def impact(self):
        return "Production AI systems processing millions of records daily"

Currently: Building AI-driven scoring systems and ETL pipelines on AWS at Arionkoder
Also: Mentoring Master's and Specialization students in AI, IoT & Embedded Systems at UBA
Background: 10 years embedded firmware (STM32) → 5+ years cloud-native AI systems

🎯 AI-Powered Stakeholder Scoring System

Automated Decision-Making & Risk Detection

Enterprise-scale reputation scoring API processing multiple data sources with integrated ML for client risk assessment and anomaly detection. Used for automated decision-making across stakeholders.

Impact: Improved risk detection and data-driven insights, real-time scoring at scale

Python FastAPI AWS Machine Learning DynamoDB Redshift GraphQL

⚡ AWS Glue Local Development Framework

Cloud-Independent ETL Testing

Local execution environment for AWS Glue pipelines that eliminates cloud dependencies during development. Mocks AWS services, translates S3 paths, provides JDBC connectivity for Redshift sources.

Impact: 60% reduction in development cycle time

AWS Glue PySpark Python Redshift MongoDB Spark Docker

🤖 AI Call Center System

Real-time Conversational Analysis

Real-time audio processing with Speech-to-Text/Text-to-Speech and LLM-powered conversational analysis. Production system handling enterprise call volumes.

Impact: 35% reduction in call handling time

Python FastAPI AWS Lambda Whisper LLMs S3 DynamoDB

🔊 GPU-Accelerated Speaker Diarization

Neural Audio Analysis Server

Sophisticated call analysis server with AI-powered speech recognition and neural speaker diarization using PyTorch and CUDA acceleration. Integrates Google Speech-to-Text with custom-trained models for high-performance speaker segmentation.

Impact: 90%+ speaker identification accuracy, on-premise GPU acceleration

FastAPI PyTorch CUDA Python GPU Optimization Audio ML

📊 Data Pipeline & Scoring Engine

ETL + NoSQL Transformation System

End-to-end pipeline extracting data from Redshift and transforming to structured NoSQL documents. Handles data seeding, parameter configuration, and metadata optimization for reputation scoring calculations.

Impact: Processes large datasets with automated score computation and dynamic parameter adjustment

FastAPI Python Redshift MongoDB ETL Data Engineering

📞 Political Campaign Analytics Platform

Call Intelligence System

Comprehensive call analytics processing thousands of daily calls with automated transcription, sentiment analysis, and geographic distribution tracking.

Impact: 75% reduction in analysis time, 40% increase in response rates

Python Streamlit Twilio AWS DynamoDB NLP Data Analysis

🏠 AI Real Estate Communication Platform

Intelligent Property Management

Automated property inquiry system with Twilio integration, real-time AI communication, and Kafka-based pipelines for managing property listings and virtual tours.

Impact: Processed hundreds of daily inquiries with high availability

Node.js AWS Amplify Twilio Kafka Kubernetes GraphQL

💳 Fintech Coupon Processing System

High-Performance Transaction Engine

Enterprise coupon handling system enabling near-instant redemption, handling 10K+ daily transactions with multi-channel payment processing.

Impact: 800ms → 200ms API response time, 45% reduction in database queries

Python FastAPI Django GCP Cloud Run PostgreSQL Redis Stripe

🔬 Holographic Data Compression

CUDA-Accelerated Algorithms

CUDA-accelerated compression algorithms for holographic data, presented at HDC Eclipse Workshop (Harrisburg, USA). Advanced compression for medical imaging and VR applications.

Impact: 70% size reduction without quality loss

CUDA C++ GPU Computing Parallel Processing AGI Safety

🎵 Meditative Music Generation

VAE-Based Generative AI

Generative AI system using Variational Autoencoders for ambient music synthesis. Built with TensorFlow for learning and replicating complex musical patterns with spectrogram analysis.

Impact: Interactive interface for personalized meditative music generation

Python TensorFlow VAE Librosa MIDIUtil Music21 Audio Processing

🎬 Video Encoding Backend

AI-Enhanced Media Processing

FFmpeg-based system with ML/AI integration for automated video/audio content processing and encoding.

Python FastAPI C++ FFmpeg OpenCV HLS/DASH

📋 View complete project portfolio on LinkedIn →

Postgraduate Studies (In Progress) — Data Engineering & Artificial Intelligence | Instituto Balseiro, 2026-Present
Focus: ML Pipelines, Deep Learning, LLMs, RAG Frameworks, Distributed Computing, PySpark, CUDA/GPU Acceleration, MLOps

Master's Degree — Internet of Things | University of Buenos Aires, 2022
Postgraduate — Embedded Systems | University of Buenos Aires, 2020
Bachelor's Degree — Automation & Robotics | University of Lomas de Zamora, 2013
Engineering Degree — Automation & Robotics | Universidad Tecnológica Nacional, 2013

Graduate Mentor & Thesis Advisor | University of Buenos Aires
Mentoring Master's and Specialization students in:

• Artificial Intelligence & Machine Learning
• Internet of Things (IoT)
• Embedded Systems & Firmware Development
• Cloud Architecture & Scalability

Member of thesis evaluation committees for graduate programs.

// Exploring Go for high-performance cloud-native systems
package main

import "fmt"

func main() {
    focus := []string{
        "Go for microservices architecture",
        "Advanced RAG implementations",
        "Real-time AI inference optimization",
        "Kubernetes at enterprise scale",
    }
    
    for _, topic := range focus {
        fmt.Printf("Learning: %s\n", topic)
    }
}