For employers

Hire this AI Trainer

Sign in or create an account to invite AI Trainers to your job.

LoginCreate account
Invite to Job
G
Giovanni Pasqualini

Giovanni Pasqualini

AI-powered Document Data Extraction (Traveloo Project)

Italy flagSan Miniato, Italy
$30.00/hrIntermediateHivemindProdigyRemotasks

Key Skills

Software

HiveMindHiveMind
ProdigyProdigy
RemotasksRemotasks
OpenCV AI Kit (OAK)OpenCV AI Kit (OAK)
MindriftMindrift
AWS SageMakerAWS SageMaker
Axiom AI
ClickworkerClickworker

Top Subject Matter

Flight booking confirmations
airline data extraction
travel document automation

Top Data Types

DocumentDocument
ImageImage
TextText

Top Task Types

Text GenerationText Generation
ClassificationClassification

Freelancer Overview

AI-powered Document Data Extraction (Traveloo Project). Brings 6+ years of professional experience across complex professional workflows, research, and quality-focused execution. Core strengths include OpenAI API, Anthropic Claude API, and Internal. Education includes Bachelor of Arts, University of Pisa (2023). AI-training focus includes data types such as Document and Image and labeling workflows including Text Generation and Classification.

IntermediateEnglishDutchItalianSpanishFrench

Labeling Experience

AI-powered Document Data Extraction (Traveloo Project)

DocumentText Generation
Created AI-powered document processing pipelines focused on extracting structured information from unstructured flight-related PDF documents. Designed and engineered prompt templates to convert free-form booking confirmations into well-formatted JSON using LLMs. Leveraged OpenAI API and Anthropic Claude API for advanced information extraction and automated entity recognition. • Developed robust LLM prompts to ensure consistent and reliable extraction. • Validated and structured airline and flight booking data for downstream processing. • Evaluated and improved model outputs against ground truth data. • Integrated outputs into interactive visualization tools for user consumption.

Created AI-powered document processing pipelines focused on extracting structured information from unstructured flight-related PDF documents. Designed and engineered prompt templates to convert free-form booking confirmations into well-formatted JSON using LLMs. Leveraged OpenAI API and Anthropic Claude API for advanced information extraction and automated entity recognition. • Developed robust LLM prompts to ensure consistent and reliable extraction. • Validated and structured airline and flight booking data for downstream processing. • Evaluated and improved model outputs against ground truth data. • Integrated outputs into interactive visualization tools for user consumption.

2023 - 2024

Coin Image Embedding & Classification (Coin Retrieval Engine)

ImageClassification
Engineered and deployed an image retrieval engine that utilizes deep embeddings and metric learning for the classification and organization of coin images. Applied ResNet-based feature extraction, followed by vector-based similarity search using FAISS to assign classes to new images. Evaluated model performance on large-scale coin datasets with fine-tuned retrieval and class assignment techniques. • Implemented metric learning with triplet loss for robust visual similarity assessment. • Labeled and classified coin images based on learned embeddings and evaluation datasets. • Optimized the system to support scalable class addition without full retraining. • Conducted recall evaluations to measure and improve annotation quality and accuracy.

Engineered and deployed an image retrieval engine that utilizes deep embeddings and metric learning for the classification and organization of coin images. Applied ResNet-based feature extraction, followed by vector-based similarity search using FAISS to assign classes to new images. Evaluated model performance on large-scale coin datasets with fine-tuned retrieval and class assignment techniques. • Implemented metric learning with triplet loss for robust visual similarity assessment. • Labeled and classified coin images based on learned embeddings and evaluation datasets. • Optimized the system to support scalable class addition without full retraining. • Conducted recall evaluations to measure and improve annotation quality and accuracy.

2023 - 2023

Education

U

University of Pisa

Bachelor of Arts, Digital Humanities

Bachelor of Arts
2023

Work History

G

Genio Diligence

Python Developer

San Miniato
2025 - Present
G

Gruppo Insieme

Backend Developer

San Miniato
2025 - 2025