Curriculum Vitae Automation

Automating C.V. read and information extraction

I developed an efficient pipeline that streamlines the extraction of key information from resumes, saving staffing companies countless hours of manual data entry. This powerful software automatically parses candidate CVs and transforms them into a standardized format, ensuring quick, seamless, and accurate processing. Perfect for staffing agencies looking to accelerate recruitment processes and improve productivity.

• Client: Particular
• Date: Nov 2024
• Service: OCR-NLP

(left) output of the pipeline, (right) C.V. to extract information

Close

Personalized Product Detector

Advancing Shelf Auditing with a Custom Product Classifier

I am actively developing a robust AI solution that leverages deep learning to enhance product detection on retail shelves. By collecting and labeling data to improve existing public datasets like SKU110k, this project aims to create a highly accurate, scalable detector for real-time product detection and future recognition for shelf auditing. The enhanced dataset will drive better performance in inventory management and reduce discrepancies, ultimately optimizing retail operations with precise, AI-powered auditing solutions.

• Client: TIA
• Date: August 2024
• Service: AI Pipeline Developement

Preliminar results of model transfern learning.

Pipeline of the detection and classification process

Close

Self-Checkout Fruits and Vegetables

Helping clients to select products faster and easier

I successfully deployed an advanced Fruits & Vegetables Classifier, originally developed by an AI lab, into production using FastAPI and MLOps techniques. With over 90% accuracy, this AI model enhances the customer experience by enabling faster and more efficient product selection. The system automatically identifies products through real-time image classification, presenting users with a dynamic screen displaying possible matches, which streamlines the selection process. This solution not only improves checkout speed but also reduces human error, offering a seamless and intuitive interaction for customers in retail environments. The integration of MLOps ensures continuous monitoring, scalability, and performance optimization, making the solution robust and future-proof for high-traffic environments.

• Client: TIA
• Date: April 2024
• Service: Client Satisfaction

Close

Client Traffic Metering

Reporting Client Traffic Metrics Seamlessly

successfully integrated innovative solutions that deliver comprehensive client traffic insights, helping businesses enhance their operational efficiency. These projects empower retailers and commercial spaces to make data-driven decisions by leveraging advanced analytics and AI-powered tools. All projects were deployed using DevOps metodologies, in a single server in the edge (store) to minimize bandwhich requirements.

• Client: TIA
• Date: February 2024
• Service: Artificial Vision Automation and Reporting

Real-time detection and management of crowd levels to ensure safety and optimal flow.

Visualizing high-traffic areas and customer movement patterns for optimized store layouts.

Close

POS Recomendation System

Recomending products in the point-of-sale based on the cart

Product Recommendation System for Point-of-Sale (POS) that enhances customer experience by providing personalized product suggestions based on the items in their cart. This AI-driven solution analyzes shopping patterns in real-time to offer targeted product recommendations directly at the checkout, increasing upsell opportunities and boosting revenue. Seamlessly integrated into POS systems, it helps retailers maximize sales while delivering tailored, relevant suggestions to customers, ensuring a more engaging and efficient checkout process.

• Client: TIA
• Date: December 2023
• Service: Client Satisfaction

Close

Obsolete Prices Detector

Reducing Time Consuming Activies

I developed and deployed an AI-driven solution for detecting and reading price tag labels directly from retail shelves using Convolutional Neural Networks (CNNs). This cutting-edge system accurately identifies and extracts pricing information from product labels, streamlining inventory audits and ensuring price accuracy in real-time. By automating the detection process with advanced AI technology, the solution reduces manual labor, minimizes human error, and enhances operational efficiency in retail environments. This innovative approach empowers retailers to maintain accurate pricing and optimize shelf management, ultimately improving the overall customer shopping experience.

• Client: TIA
• Date: July 2022
• Service: AI Pipeline Developement

Process of personalized detection of price tag labels.

A snapshot of the reporting platform, showcasing alerts for pricing discrepancies and obsolete items.

A close-up view of a real alert, demonstrating the system's accuracy and immediacy in addressing issues.

Close

CCTV Anomalies Detection

Automated Network Anomalies Detection for a Retail CCTV

Automated Network Anomalies Detection for Retail CCTV seamlessly integrates with Hikvision SDK to proactively identify and resolve network device issues. This advanced solution detects connection errors, time synchronization problems, and other anomalies, ensuring uninterrupted surveillance and optimized performance. It enhances security operations by simplifying the identification of faulty network devices, reducing downtime, and improving overall system reliability. It was designed for retail, banks, or any environment with surveillance needs.

• Client: TIA
• Date: September 2020
• Service: Device Integration

Close

Lucia

Revolutionizing Retail with Autonomous Shelf Auditing

Cutting-edge Autonomous Robotic System engineered to deliver precise and efficient shelf auditing in dynamic retail environments. By harnessing the power of AI and Compute Vision, LUCIA performs real-time shelf analysis, significantly enhancing inventory management while reducing human error and operational costs. LUCIA is a transformative solution for retailers, providing actionable insights that streamline operations and improve overall store performance.

• Client: TIA
• Date: February 2019
• Service: Robotic Automation

3D design renderings of LUCIA and its charging station.

The fully functional robot in action, safeguarded within a protective enclosure in a live store environment.

Close