Diogo Dutra, Developer in Toronto, ON, Canada
Diogo is available for hire
Hire Diogo

Diogo Dutra

Verified Expert  in Engineering

Bio

Diogo is a machine learning engineer with a Master of Science degree in aeronautics and expertise in the whole pipeline from data visualization to model deployment. His 14+ years of international experience in research and development for defense, public safety, and aerospace industries allow him to combine new methodologies, like deep learning and computer vision, with traditional ones (e.g., physics, simulation, sensor fusion, Kalman filter, Fourier transform, and signal processing).

Portfolio

Ford Motor Company
Machine Learning, Computer Vision, Deep Learning, ARM, Android, AUTOSAR...
Patriot One Technologies
Data Science, Data Analysis, Image Processing, Technical Leadership...
Altran
Agile, Google Test, CMake, AUTOSAR, Visual Studio Code (VS Code), OpenCV, Git...

Experience

  • Digital Signal Processing - 7 years
  • Research - 5 years
  • Machine Learning - 3 years
  • Computer Vision - 3 years
  • PyTorch - 3 years
  • Python 3 - 3 years
  • Data Science - 3 years
  • Deep Learning - 2 years

Availability

Part-time

Preferred Environment

Python 3, SciPy, Scikit-learn, Pandas, PyTorch, Jupyter Notebook, Fast.ai, Aerospace & Defense, Digital Signal Processing, Computer Vision

The most amazing...

...thing I've developed detects and identifies concealed weapons (e.g., handguns) and non-weapons (e.g., phones) on the body by combining deep learning and radar.

Work Experience

Machine Learning Developer

2021 - PRESENT
Ford Motor Company
  • Tasked with the research and development of future artificial intelligence applications for the next generation of autonomous vehicle infotainment systems at Ford's R&D laboratory.
  • Conceived a series of PoCs for the self-driving car, including different areas, such as deep learning, computer vision, user recommendation, embedded systems, and connectivity with other devices.
  • Utilized many technologies, including Linux, Office365, Eclipse, Blackberry QNX, CMake, C++, Android Auto, AUTOSAR, Python, OpenCV, TensorFlow, Git, and Google Cloud Platform.
Technologies: Machine Learning, Computer Vision, Deep Learning, ARM, Android, AUTOSAR, Android Auto, Google Cloud Platform (GCP), TensorFlow, Linux, QNX, Git, Python, Python 3, Python 2, Office 365, C++, CMake, Ubuntu, Windows, Docker, FastCV, OpenVX, Docker Compose, Windows Subsystem for Linux (WSL), Object Detection, Image Segmentation, Biometrics, Facial Recognition, Augmented Reality (AR), Snapdragon Neural Processing Engine (SNPE), Mathematics, OpenCV, Open Neural Network Exchange (ONNX), Development, PyTorch, TensorFlow Lite, Bash Script, Technical Hiring, Code Review, Source Code Review, Interviewing, Computer Vision Algorithms, 3D, Object Recognition, You Only Look Once (YOLO), Algorithms

Lead Data Scientist

2019 - 2021
Patriot One Technologies
  • Created a deep learning model that detects concealed weapons with radar and magnetic sensors, digital signal processing, and a convolutional neural network.
  • Managed a technical team consisting of a principal scientist and data scientists on an Agile project.
  • Accumulated management roles as machine learning engineer and data science team leader with my original role as a data scientist.
Technologies: Data Science, Data Analysis, Image Processing, Technical Leadership, Product Development, Research, Supervised Learning, Neural Networks, Convolutional Neural Networks (CNNs), Data Cleaning, Data Visualization, Data Mining, Fourier Transform (FT), Python, Confluence, Jira, Network Analysis, Radar, Seaborn, Pandas, Scikit-learn, SciPy, PyTorch, Python 3, Anaconda, Jupyter Notebook, Git, Machine Learning, Team Leadership, Teamwork, Predictive Analytics, Predictive Learning, Object Detection, Digital Signal Processing, Statistical Modeling, Docker, Visualization, Signal Processing, Autoencoders, Proof of Concept (POC), Mathematics, Linear Regression, OpenCV, Technical Hiring, Interviewing, Code Review, Source Code Review, 3D Image Processing, Computer Vision Algorithms, 3D, Object Recognition, Semantic Segmentation, Algorithms, Principal Component Analysis (PCA), Correlational Analysis, AI Consulting

Senior Consultant Engineer

2018 - 2019
Altran
  • Coded embedded C++ for the BMW and Continental future fully digital driver's dashboard, specifically the BMW Series 7 Instrument Cluster 5th Generation.
  • Updated procedures on the Confluence website such as onboarding instructions for new team members and how to set up the environment.
  • Applied Google unit tests and Python Robot framework to automate testing procedures for verification of new versions before release.
Technologies: Agile, Google Test, CMake, AUTOSAR, Visual Studio Code (VS Code), OpenCV, Git, Continuous Integration (CI), Confluence, Jira, ARM, C++, Python 3, Robot Framework, Teamwork, Python, Algorithms

Technology Consultant

2017 - 2018
Natura Brazil
  • Completed a data science analysis of an MRO warehouse, reducing 30% of the spare parts stock value without impacting stock-out.
  • Promoted industry 4.0 applications, including additive manufacturing (3D printing), augmented reality, IoT, big data, machine learning, and artificial intelligence.
  • Presented a successful business case for acquiring a professional 3D printer that reduced the acquisition of plastic SKUs by 66% annually.
  • Implemented three proofs of concept with augmented reality.
  • Served as the scrum master for developing and employing a new system to request materials and services. Increased the administrative assistant team's productivity by 48% and improved access to information for the internal maintenance team.
Technologies: SAP, Predictive Maintenance, 3D Printing, Augmented Reality (AR), Office 365, Microsoft Power BI, Azure, Leadership, Team Leadership, Agile Leadership, Python, Industry 4.0, Predictive Modeling, Warehouses, Forecasting, Technical Leadership, Management, Predictive Analytics, Predictive Learning, Python 3, Visualization, Mathematics, Dashboards, Mobile Apps, Interviewing, Algorithms, AI Consulting

Chief of Technology Office

2016 - 2017
GoEpik
  • Turned my startup GoEpik into the most attractive Brazilian startup by 2017 while I was the CTO by offering augmented reality for field maintenance. We were accelerated by Google and Plug and Play with an invitation to stay in Silicon Valley.
  • Conceived and coded the first prototype to visually guide the user to record proofs of task completions with pictures and ease maintenance activities, such as inspection, cleaning, and lubrication. It was converted into sales for Natura and Renault.
  • Conceived and coded the second prototype to connect an expert with a field technician through video conference, using an augmented reality platform with computer vision features such as adding markers, texts, and pictures in the field of view.
  • Benchmarked similar products and competitors for marketing positioning, including SightCall, Microsoft HoloLens, and Scope AR.
Technologies: RESTful Development, REST, CTO, WebRTC, Airtable, Vuforia, Android, C#, Unity3D, REST APIs, Technical Leadership, Minimum Viable Product (MVP), Proof of Concept (POC), Mathematics, Algorithms

Combat System Integration Engineer

2013 - 2017
Itaguai Construcoes Navais
  • Performed assembly and inspection specifications (AIS); setting-to-work (STW); factory acceptance tests (FAT); and integration, verification, validation, and qualification (IVVQ) activities.
  • Participated in the transfer of technology at former DCNS premises for the combat system integration activities during on-the-job training (OJT) under the new Riachuelo class (derived from the French Scorpène) for the Brazilian submarine program.
  • Translated technical documents from English and French to Portuguese.
Technologies: IP Networks, Windows, Wireshark, HP Quality Center (QC), Translation, Project Management, PeopleTools, Requirements, Digital Signal Processing, Technical Leadership

Project and Commercial Manager

2012 - 2013
Mectron
  • Prospected a R$200,000 new contract for embedded software development from DCNS, a French Naval and Defense exporter, by working all the way from first cold contact with the client until the contract signature.
  • Led engineers who were expatriated to France for software development on the client’s premises.
  • Accumulated the responsibility for the contract prospected by me. Embedded software development for PLC under the PMI methodology to French clients.
  • Reported my project progress to stakeholders, including the CEO, the new business director, and the French client DCNS.
  • Finished the project within the deadline, and its profit was above expectations, from a loss to 10% positive.
  • Wrote contracts, including pricing definitions and answers to requests for information and proposals (RFI/RFQ).
  • Represented as an exhibitor in many international defense events, such as the Brazilian LAAD and IDEX in the United Arab Emirates.
Technologies: Leadership, Team Leadership, Aeronautics, Simulations, Navigation, Sensor Fusion, Kalman Filtering, Inertial Navigation System (INS), GPS, MATLAB, Simulink, C, C++, CMake, ARM, Project Management, PMBOK, Aerospace & Defense, Aerodynamics, Telemetry, Pricing, Contract, Business to Business (B2B), Physics Simulations, Controls, Unmanned Aerial Vehicles (UAV), Sensor Data, GPSS, Digital Signal Processing, Technical Leadership, Interviewing

Aerospace Engineer

2007 - 2013
Denel Dynamics
  • Developed real-time data fusion for missiles and guided bombs: transfer alignment for initialization of inertial platform and aided navigation (GPS, BaroAltimeter, and aircraft INS).
  • Validated the embedded code on hardware-in-the-loop simulation, embedded on ARM using C language, and analyzed flight test data by telemetry in the field at Air Force launching base.
  • Increased the flight span duration from 1 to 10 minutes thanks to better navigation accuracy.
  • Created a novel target tracking algorithm for missile employing extended Kalman filter by performing data fusion of passive target line-of-sight signals based on 6-states extended Kalman filter with design in MATLAB and embedded in C code.
  • Automated code unit tests of in-house linear algebra and Kalman filter libraries.
  • Simulated missiles subsystems including moving parts, hydraulic, external, and internal aerodynamics (subsonic and supersonic) on Simulink.
  • Defined the technical specification of the inertial unit (gyroscope and accelerometer) for a long-range missile.
  • Translated technical documents such as textbooks, papers, and technical reports from English to Portuguese.
Technologies: C++, C, Cloud Data Fusion, Kalman Filtering, Inertial Navigation System (INS), Aeronautics, Unmanned Aerial Vehicles (UAV), Aerospace & Defense, Aerodynamics, Autopilot, Controls, Navigation, ARM, Simulations, Hardware-in-the-loop (HIL), Simulink, MATLAB, Estimators, Embedded C, Teamwork, Digital Signal Processing, Signal Processing, Mathematics, 3D, Algorithms

Aerospace Engineer

2007 - 2008
Embraer
  • Retrofitted the propulsion system for the F5-BR jet fighter and AMX (A-1M) jet bomber, including field tests with the airplane anchored to the ground and its turbine operating with afterburner on.
  • Created a simulation of the Rolls-Royce Spey Mark 807 engine in MATLAB to support other departments such as aerodynamics.
  • Managed propulsion system project requirements with DOORS software.
Technologies: Datacom, DOORS, Simulink, MATLAB, Aerodynamics, Aeronautics, Aerospace & Defense, Teamwork, Simulations, Digital Signal Processing, Signal Processing, Mathematics, 3D, Algorithms

Fake Face Generator

https://arxiv.org/pdf/1511.06434.pdf
Trained generative adversarial networks (GAN) using a dataset of celebrities' photos to learn how to create a fake human face. This same technique can generate real content like images, texts, simulations, or signals.

Covid-19 Triage by Hemogram

Created this classifier to help hospitals triage the suspects of COVID-19 from a simple blood test! It is faster and spares many SARS-CoV-2 test kits. It can save lives and costs simultaneously during the peak of the coronavirus pandemic.

Diagnose Pneumonia

What if we could speed up the diagnosis process? A novel classifier was trained with a dataset of X-ray images from children's chests. Now, it labels any x-ray image as either normal, bacterial pneumonia, or viral pneumonia. The result is an automated diagnosis that can accelerate the clinical care for children with world-class precision.

Concealed Weapon Detector

https://patriot1tech.com/
Applied computer vision and digital signal processing techniques on the radar for classification. The novel model finds meaningful contrasts between weapons and decoys of a person in movement. CNN can be used to analyze images and any other signals such as video, audio, radar, and sonar.

AI Writer

What if artificial intelligence could kickstart an original text, so you do not have to start a new one from scratch? I created this website that the user enters a few keywords, and it gives back some related suggestions of text with titles. I used Heroku, RESTful, and asynchronous tasks to deploy this model online.

Online Product Recommender for eCommerce

Based on a collaborative filter, I created a recommendation engine for a cloud-service online retailer. Any online business must leverage its database to find out and deploy a better user experience to drive consumption. Unsupervised learning is ideal to find customer segments, which can be used to predict common preferences.

Flight Delay Predictor

The world's busiest airways passengers now know if their flight will delay with 15 minutes of margin. I created this supervised learning model, considering a 2-year long hourly dataset of airport departures, arrivals, and weather.

Quadcopter Autopilot

https://arxiv.org/pdf/1509.02971
The Parrot quadcopter just gained an autopilot. Thanks to thousands of simulated runs, this new reinforcement learning knows how much power each rotor must apply to stabilize the drone. Let me know if you need specific maneuvers or autopilot for any other vehicle.

Bike Sharing Demand Forecast

Now, the owner of a bike-sharing company from Washington can make better decisions on how many bikes to buy. This forecast helps the owner to avoid losing revenue by having too few or skyrocketing the operational costs by having too many. I created this model by training an RNN over a 2-year hourly rental and weather data. RNNs are suitable to predict any time-series data such as demand, weather, and stock options.

Retailer Customer Segmentation

A retailer from Portugal wanted to lower the frequency of deliveries to reduce costs without considerably disturbing its clients. It was achieved by finding customer segments and selecting which would not be significantly impacted by three weekly deliveries instead of five. Unsupervised learning is ideal for finding customer segments and understanding their specific purchase behaviors.

Painting Style Copier

The style transfer technique allows the convolutional neural network to learn the content independently from one picture and the style from another. Later, I gradually applied the style from the painting to the content of the picture with the octopus.

Missile Autopilot (Navigation, Guidance, and Command)

Designed, embedded, and tested in the field with telemetry during real launches algorithms for missiles and smart-bombs, including aided navigation (GPS, BaroAltimeter, and IMU), transfer alignment, sensor fusion, target position estimation, and extended Kalman filter.

Sentiment Analysis for Movie Reviews

Created a predictive model to read a movie review and classify it as POSITIVE or NEGATIVE. The model applies text processing techniques and a neural network. This technique is also reusable for other applications such as product reviews or stock market discussions.

Brandify, Online Addition of Logo on Background of an Awesome Ad Image

Online advertisement opens room for custom experience based on the user's known preferences. This opportunity is explored by offering an automated way to change the image on website banners with the brand based on which user is accessing the website.

This methodology is achievable by reusing a trained ResNet101 (transfer learning) for image segmentation, followed by image processing to find the best spot on the background to place the logo.

Virtual Try-on (Virtual Showroom, Augmented Reality with Real Garments)

Structured the technologies for a new virtual showroom startup. I planned the scope of the proof-of-concept (POC) and minimum viable product (MVP) for a startup about to launch a new mobile shopping experience by combining augmented reality, photogrammetry, artificial intelligence, simulations, and virtual try-on.

The outputs are a photorealistic image of the user wearing the virtual garment and a heatmap as a realistic representation of how the garment will fit and the proper size for the specific user.

I also proposed new state-of-the-art solutions (better than photogrammetry) to scan the human body accurately.

Video Special Effects (Instagram-style Filters)

https://youtu.be/Q1P3Hq6fsoE
I developed the video processing technology core for one of the big five giant American tech companies. Now you can use your camera to create custom special effects that rely on background removal and augmented reality animations. This technique is transferable to online streaming data, such as filters.

Data Science for Forex

A series of studies to find opportunities for Forex trading, including:
• Analysis and feature engineering of historical real trading data.
• A custom-trained model with emphasis on explicability.
• A custom Kalman Filter for intraday.
• Time-series ARIMA analyses.
• Autocorrelation.
• Pair trading.

Human Motion Algorithms

I created an algorithm to detect human body posture from wearable inertial sensors using sensor fusion with Kalman filter and forward kinematics. The methodology was implemented using state-of-the-art paper. Results were compared against MOCAP (motion capture) data.

Face 3D Scanning and Reconstruction with Texture

https://youtu.be/Tvfz1m3d0Og?si=Kzgq5wn6HBW8qJr6
3D scanning/reconstruction of a human face from two selfies: one frontal and one lateral. The reconstruction output includes the 3D cloud points and the image texture. This work was commissioned by a company in the virtual reality gaming industry, but it can also have spin-offs for other industries.
2022 - 2022

Master's Degree in Computer Science

Georgia Tech - Atlanta, Georgia, USA

2011 - 2013

MBA in Entrepreneurship

Getulio Vargas Foundation (FGV) - Rio de Janeiro, Brazil

2006 - 2010

Master's Degree in Aerospace Engineering

Institute of Aeronautical Technology - São José dos Campos, Brazil

2001 - 2005

Bachelor's Degree in Mechanical Engineering

Institute of Military Engineering - Rio de Janeiro, Brazil

JUNE 2020 - PRESENT

AI Product Manager

Udacity

JUNE 2019 - PRESENT

Deep Learning

Udacity

JUNE 2018 - PRESENT

Machine Learning

Udacity

DECEMBER 2017 - PRESENT

Data Science Essentials

edX

JUNE 2017 - PRESENT

Translator English Portuguese

DBB

Libraries/APIs

PyTorch, Pandas, Scikit-learn, NumPy, Matplotlib, SciPy, OpenCV, FFmpeg, WebRTC, REST APIs, Beautiful Soup, Fast.ai, TensorFlow, OpenVX, Keras, Asyncio, Blender Scripting API

Tools

Seaborn, Radar, Microsoft Power BI, Hardware-in-the-loop (HIL), Blender, MATLAB, Git, AutoML, Jira, OpenAI Gym, Confluence, CMake, HP Quality Center (QC), Wireshark, DOORS, Composer, GitHub, PeopleTools, Android Auto, Docker Compose, Snapdragon Neural Processing Engine (SNPE), ARIMA, Open Neural Network Exchange (ONNX), You Only Look Once (YOLO)

Languages

Python 3, C++, Python, C, SQL, Simulink, C#, GPSS, Embedded C, Python 2, Bash Script, Bash

Frameworks

Flight, Unity3D, Google Test, Flask, Robot Framework, Robot, TensorFlow Lite, MediaPipe, Qt

Platforms

Jupyter Notebook, Anaconda, Windows, Heroku, Azure, Linux, Visual Studio Code (VS Code), Android, Vuforia, Docker, Amazon Web Services (AWS), Google Cloud Platform (GCP), Ubuntu, Particle, Oculus

Paradigms

Management, Agile, REST, Continuous Integration (CI), PMBOK, RESTful Development

Industry Expertise

Project Management, Marketing

Storage

Datacom, SQLite

Other

Digital Signal Processing, Kalman Filtering, Aerospace & Defense, Aeronautics, Supervised Learning, Machine Learning, Predictive Modeling, Regression, Data Analysis, Deep Learning, Convolutional Neural Networks (CNNs), Autoencoders, Learning Transfer, Metrics, Translation, Neural Networks, Artificial Intelligence (AI), Computer Vision, Classification, Deep Neural Networks (DNNs), Product Development, Analysis, Trend Analysis, Image Processing, Navigation, Autopilot, Unmanned Aerial Vehicles (UAV), Cloud Data Fusion, Fourier Transform (FT), Signal Processing, Data Science, Visualization, Statistical Modeling, Research, Engineering, Image Recognition, Sensor Fusion, Inertial Navigation System (INS), Decision Tree Classification, Decision Tree Regression, Decision Trees, Linear Regression, Polynomial Regression, Mathematics, Physics, 3D, Computer Vision Algorithms, Object Recognition, Algorithms, Principal Component Analysis (PCA), Correlational Analysis, AI Consulting, Simulations, Controls, Entrepreneurship, Unsupervised Learning, Clustering, Generative Adversarial Networks (GANs), Writing & Editing, GPU Computing, Graphics Processing Unit (GPU), Time Series, Product Forecasts, Sales Forecasting, Augmented Reality (AR), Statistics, Data Analytics, CTO, Logistic Regression, Numerical Simulations, Consulting, 3D Image Processing, 3D Rendering, Aerodynamics, Aircraft Engineering, Finance, Administration, Projects, Reinforcement Learning, Data Inference, Analytics, Deployment, Natural Language Processing (NLP), Trend Forecasting, Time Series Analysis, Recurrent Neural Networks (RNNs), Product Lifecycle Management (PLM), Business Cases, Annotations, Grammar & Language Creation, Texting, APIs, Kaggle, Online Sales, eCommerce, Markov Model, Deep Reinforcement Learning, Network Analysis, Office 365, 3D Printing, Predictive Maintenance, SAP, Airtable, IP Networks, Bokeh, Data Visualization, Data Mining, Data Cleaning, Technical Leadership, Web Scraping, Scraping, Recommendation Systems, Text Processing, Text Classification, Leadership, Team Leadership, Agile Leadership, GPS, ARM, Telemetry, Industry 4.0, Warehouses, Forecasting, Requirements, Teamwork, Pricing, Contract, Business to Business (B2B), Physics Simulations, Sensor Data, Estimators, Predictive Analytics, Predictive Learning, Predictive Text, Image Segmentation, AUTOSAR, QNX, MPEG, Scripting, Videos, Object Detection, Articles, Planning, MVP Design, Minimum Viable Product (MVP), Proof of Concept (POC), Lean Startups, Startups, Virtual Reality (VR), Photogrammetry, Fashion, PNG, MP4, Video Processing, FastCV, Windows Subsystem for Linux (WSL), Robotics, Simultaneous Localization & Mapping (SLAM), Localization, Search, Mapping, Monte Carlo, Policy, PID Controllers, Simulators, White Papers, Biometrics, Facial Recognition, Linear Algebra, Forex Analysis, Development, Trading, Stock Trading, Stock Market, Dashboards, Mobile Apps, Technical Hiring, Code Review, Source Code Review, Interviewing, Data-informed Recommendations, Generative Pre-trained Transformers (GPT), Statistical Analysis, Motion Capture, Wearables, Wearable Technology, Spatial Analysis, Leap Motion, Inertial Measurement Unit, Semantic Segmentation, AI Research, SWOT Analysis, Discovery, Point Clouds, Point Cloud Data, Cloud Point, 3D Reconstruction, 3D Images, Facial Landmarks, 3D Textures, 3D Texturing, 3D Scanning, Transcreation, Multi-agent Systems

Collaboration That Works

How to Work with Toptal

Toptal matches you directly with global industry experts from our network in hours—not weeks or months.

1

Share your needs

Discuss your requirements and refine your scope in a call with a Toptal domain expert.
2

Choose your talent

Get a short list of expertly matched talent within 24 hours to review, interview, and choose from.
3

Start your risk-free talent trial

Work with your chosen talent on a trial basis for up to two weeks. Pay only if you decide to hire them.

Top talent is in high demand.

Start hiring