David Lopes AI Engineer
LinkedIn

AI Engineer at SafeCaring · Coimbra

Systems that see, learn and decide.
Built for the real world.

I work on computer vision, robotics and machine learning. After hours I build my own products — McFinds and Overbench.

See projects Get in touch

Computer vision · Robotics · Machine learning · RAG

david-lopes / profile.py
SafeCaring ex-Stellantis
Portrait of David Lopes
David Lopes AI Engineer · Coimbra, PT
python close the loop — data → model → decision 
# profile.py
david = Engineer(
    role="AI Engineer",
    company="SafeCaring",
    focus=["computer vision", "ML", "robotics"],
    degree="MSc Automation @ ISEC",
    shipping=["McFinds", "Overbench"],
)

david.run(pipeline=["data", "model", "decision"])
✓ loop closed
What I build

Four areas, one loop

Sense the world, decide with models, and make the result work in the field.

  • Activity & pose

    Computer vision that recognises activities, objects and human poses — trained, fine-tuned and shipped as modular services.

  • RAG systems

    Document assistants that ground LLMs in private knowledge for accurate, context-aware answers.

  • Robotic vision

    3D cameras guiding collaborative robots in real time — like the automated nut-tightening cell I built for vehicle assembly.

  • Autonomous systems

    Navigation and line-following with OpenCV, machine learning and ROS.

Projects

Products I built from scratch

Two products of my own — designed, coded and run by me, live in production.

McFinds landing page

McFinds

mcfinds.pt

Investing research platform with an agent pipeline, citations and scorecards — markets, SEC filings and signals from video.

  • AI agents
  • RAG
  • Python
  • React
Overbench landing page

Overbench

overbench.mcfinds.pt

Desktop cockpit for coding with multiple AI models in parallel — a Director splits the task, Workers build and a Critic verifies.

  • Multi-agent
  • Desktop
  • Claude · Cursor · Codex
Journey

From the lab to production

Autonomous robots at university, car factories, and now AI in production.

Experience

  1. 01 Current

    SafeCaring

    AI Engineer

    Jul 2025 — Present

    Healthcare AI: models that recognise activities, objects and human poses — trained and fine-tuned in-house, running as modular, distributed services. Plus LLM applications.

    • Computer vision
    • Pose estimation
    • Distributed systems
    • LLMs
  2. 02 R&D

    Stellantis

    Researcher

    Jun 2023 — Jun 2025

    • Automated nut tightening. A collaborative robot with 3D vision that corrects its position in real time on the assembly line — replacing a manual step with a faster, more reliable one.
    • Robot fleet monitoring. A web platform that supervises a multi-brand fleet of mobile robots in real time, deployed in production to catch faults fast.
    • Cobots
    • 3D vision
    • AGVs
    • Web

Education

  1. 03 MSc

    MSc Electrical Engineering

    Automation · ISEC, Coimbra

    Sep 2022 — Apr 2025 · Grade 16/20

    Thesis: supervision and monitoring software for a fleet of AGVs.

    • AGVs
    • Fleet monitoring
    • Real-time
  2. 04 BSc

    BSc Electrical, Electronics & Communications Eng.

    ISEC, Coimbra

    Sep 2019 — Sep 2022

    Project: autonomous driving model for a mobile robot — vision-based line following with ML, integrated with ROS.

    • OpenCV
    • ROS
    • Python
Stack

Everyday tools

  • Pythoncore
  • PyTorchML
  • TensorFlowML
  • OpenCVvision
  • ROSrobotics
  • RAG / LLMsAI
  • FastAPIbackend
  • Reactfrontend
  • Dockerinfra
  • Linuxinfra

Want to talk vision, robots or a new product?

Open to chats about applied AI, side projects and collaborations.

LinkedIn GitHub