Building meaningful AI • Cloud pipelines
I am an Electronic Engineer focused on:
- Generative AI for scientific simulation
- Fast calorimeter simulation (CaloDiT-2, DiT, diffusion models)
My goal is to build technology with purpose.
Current:
- MSc in Data Science
- Developing cybersecurity ETL platform (Dark Eye Core)
- Studying generative models for detector simulation
Strengths:
- Deep Learning (Diffusion, DiT, Transformers)
- Big Data (Spark, distributed pipelines)
- AWS cloud-native design
- Research-oriented engineering mindsetLanguages
Python • R • SQL • MATLAB • C/C++
AI / ML
TensorFlow • PyTorch • scikit-learn • HuggingFace • XGBoost • Apache Spark
Cloud & DevOps
Docker • AWS (ECR, EKS, S3, IAM) • Kubernetes • Airflow • Git / GitHub Actions
IoT & Embedded
ESP8266 • TIVA • PSoC • Digital Signal Processing
IoT & Embedded Systems
- Ultrasonic time-of-flight sensing
- Low-cost meteorological instrumentation
- Autonomous IoT design
Docker + AWS + Microservices
Features:
- Multi-source extraction (AbuseIPDB, URLhaus, OTX)
- Normalization + cross-indicator detection
- Modular Dockerized architecture
Repo: dark_eye_core
NLP + PySpark + HuggingFace
Hybrid sentiment pipeline (VADER + transformer) • Topic modeling (LDA) • Semantic clustering (SentenceTransformers) • Named Entity Recognition
Repo: Weather-Mood-Sentiment-NLP-Analysis
Narrative AI + Conversational Memory
Story coherence using memory modules • Personality & character modeling • Modular world-state engine
Repo: Harry-Potter-LLMs-Adventure
XGBoost + Optuna + Full Pipeline
74% accuracy • ROC AUC ≈ 0.88 • Buy/Sell/Hold classifier • 15+ technical indicators • End-to-end workflow
Repo: Crypto-ML-Predictor
LinkedIn: linkedin.com/in/paulina-peralta-916a46140
GitHub: github.com/PaulinaIA
Email: pauliperalta@gmail.com
Location: Barcelona

