
Lauranrtck V
Habilidades

Conheça meus serviços


Experiência profissional
Senior iOS AI Engineer
SWIFT • Autônomo
Aug 2021 - Present • 4 yrs 9 mos
Designed and shipped multiple iOS apps leveraging Apple's LiDAR scanner for real-time 3D room mapping and floor plan generation with sub-centimetre accuracy. Built a real-time object detection pipeline using Core ML and Vision framework, achieving inference speeds under 30ms on-device without server dependency. Delivered end-to-end OCR solutions using Vision and Tesseract for document scanning, receipt parsing, and live text extraction across multiple client projects. Architected interactive sensor map UIs with ARKit and SceneKit, rendering live depth data as explorable 3D point clouds within SwiftUI applications. Reduced client onboarding friction by 60% through AI-assisted form auto-fill powered by on-device OCR and NLP classification models.
Mobile AI Developer – Spatial Mapping
LIDA360 • Período integral
Apr 2020 - May 2022 • 2 yrs 1 mo
Developed a LiDAR-based floor plan app for a property management firm, enabling agents to scan and export accurate room layouts directly from iPhone Pro in under 2 minutes. Integrated custom-trained YOLO models into a construction site safety app to detect PPE compliance in real time using the device camera. Processed and visualised dense 3D point cloud data from LiDAR sessions, exporting results in standard formats (OBJ, USDZ) for use in BIM workflows. Collaborated with product teams to translate spatial data into intuitive swipe-and-tap UX flows, reducing user learning curve from days to minutes.
iOS Developer – Computer Vision & OCR
Metadams • Freelance
Mar 2018 - Jan 2020 • 1 yr 10 mos
Built a high-accuracy invoice scanning tool using Vision framework, extracting structured fields (vendor, amount, date) with 95%+ field-level accuracy across multiple document formats. Implemented barcode and QR code scanning pipelines for a logistics tracking app, processing over 10,000 daily scans with near-zero error rate. Trained and deployed custom Core ML classifiers for document type detection, reducing manual review overhead by 40% for a FinTech client. Optimised on-device model performance using Core ML model compression, cutting memory footprint by 35% without accuracy degradation.