Lisandro A. Cesaratto

Leading development of a RAG-based internal knowledge chatbot Designing and implementing a batch tracking system for raw materials in ceramic mass preparation Created a centralized application catalog for digital production tools Developed a cloud-connected input tool for capturing production parameters with quality control and mobile QR access Headed an AI project for camera-based defect detection, from concept and hardware selection to production integration

Developed unsupervised algorithms to analyze invoice data and improve general models Upgraded a Shiny application by adding new views, reducing latency, and designing efficient API endpoints and database structures to boost user engagement

Ported a 10k+ line R fraud detection library to Python with focus on readability, runtime, and memory efficiency Integrated automated tests into GitLab CI/CD pipelines for production deployment

Supported Non-Intrusive Load Monitoring to detect device usage in home networks Built and hardware-integrated a working CNN model in Keras in under two months Designed a simulation to support project funding efforts

Built a Named Entity Recognition (NER) module for invoice data extraction with 95% accuracy, reducing tax consultant workload Onboarded and mentored 4 new team members, enhancing collaboration and productivity

Created an NLP model to automate company trustworthiness checks, cutting processing time to under 1 minute Tripled recommendation system performance through clustering and SQL-optimized Apriori algorithm

Spezialisierung: Natural Language Processing

Digitale Systeme Spezialisierung