Tiago Teixeira

Artificial Intelligence Engineer

• Enhanced autonomous aircraft landing systems using advanced computer vision and machine learning to improve navigation accuracy and safety.
• Built and deployed a modular AI pipeline for object detection, pose estimation, and segmentation using MLOps best practices.
• Optimized embedded AI models through quantization to achieve real-time performance on resource-constrained hardware.
• Developed speech-to-text prototypes and LLM-powered tools to improve human-machine interaction and accessibility.
• Collaborated with cross-functional teams in France and Silicon Valley to drive AI-based navigation innovation.

Computer Vision Intern

• Applied computer vision and ML algorithms (including RNNs and tracking models) to accurately detect accidents in dashcam videos.
• Managed and preprocessed large datasets of dashcam footage.
• Trained RNNs to recognize accident patterns, improving detection accuracy with real-world video data.
• Developed an interactive Dash application to showcase end-to-end accident detection.

Data Scientist Intern

• Developed a Multi-Agent System (MAS) simulation model to analyze smart meter communication flows and optimize network performance.
• Built interactive Dash dashboards for visualizing simulation results to aid stakeholder decision-making.
• Applied advanced data analysis techniques to improve simulation accuracy and reduce prediction error by 20%.

Front End Developer

• Developed the front-end of a SaaS platform for a cinema production team using React.
• Collaborated with a small team of freelancers to deliver a robust and user-friendly solution.

AI-Powered Invoice Analysis

Developed an AI-powered invoice analysis tool that extracts and analyzes data from invoices using Mistral OCR and LLMs.

• Extract key fields: invoice number, date, supplier, line items, VAT, totals
• Handle any format: scanned images, PDFs, handwritten notes
• Aggregate data for budgeting and expense tracking
• Detect anomalies: duplicate invoices, unusual spending patterns, missing fields
• Provide insights: suggest areas for cost savings or supplier renegotiations
• Speed up accounting by reducing manual data entry and errors

AI Radiology Assistant

This project enhances radiology workflows with a multi-agent AI system. Each agent specializes in different tasks such as image analysis and report generation

Agentic AI Writing Assistant

This project automates the editorial process with a multi-agent AI system. Each agent handles drafting, editing, and fact-checking to deliver high-quality content efficiently.

🧠 What I can build with Agentic AI:

• End-to-end content pipelines (writing, research, summarization...)
• Workflow assistants (email triage, data reporting, lead generation...)
• Automation bots (task delegation, file handling, multi-tool coordination...)
• Custom AI copilots (interactive tools with memory, reasoning, and API control...)

AI-Calendar Assistant

This project showcases the combination of text-to-speech workflow automation and speech-to-text integration connected to Google Calendar. You can also add custom automations for scheduling and reminders.

Automated Video Creation Workflow

Developed an automated video creation workflow using n8n, integrating LLMs for scriptwriting, text-to-speech for narration, and React-based tools for video editing.

AI Drone

Initiated a project on gesture recognition to control a drone.
Used OpenCV and PyTorch to build the project and implemented it to a DJI Ryse Tello Drone.

AI Hand | see project on GitHub

Created a Python application to perform mathematical operations using sign language.
Processed information about the user’s signs to solve an equation.
Utilized OpenCV for image processing, a Deep Learning model and Kaggle competition dataset.

AI CAR | see project on GitHub

Designed a virtual autonomous car that learns to drive on a circuit in Python.

CentraleSupélec