Carl Dehlin, Developer in Stockholm, Sweden
Carl is available for hire
Hire Carl

Carl Dehlin

Verified Expert  in Engineering

Algorithm Developer

Stockholm, Sweden
Toptal Member Since
August 23, 2021

Carl is a mid-senior algorithm developer with experience in optimization and artificial intelligence, primarily for applications in vision such as 3D reconstruction and object detection. Currently, he is working on a startup project on speech recognition. Carl codes mainly in C++ and knows the language inside out. As an expert in his field, he can easily code in new languages and find intelligent solutions for complex problems.


C++, 3D Graphics, Artificial Intelligence (AI), Bayesian Statistics...
Cubist IT
C++, Artificial Intelligence (AI), Bayesian Statistics, Computer Vision, Linux...
MATLAB, Python, C, Optimization, Machine Learning, Neural Networks...




Preferred Environment


The most amazing...

...thing I have developed is an automatic trailer tracker algorithm used as part of a smart mirror system.

Work Experience

Computer Vision Engineer

2020 - 2021
  • Developed a multi-sensor SLAM using LiDAR, camera, wheel odometry, IMU, and GPS.
  • Created a multi-rate filter for estimating a signal without making unnecessary jumps.
  • Worked with a map matching algorithm used for snapping vehicle trajectories to a road network.
Technologies: C++, 3D Graphics, Artificial Intelligence (AI), Bayesian Statistics, Computer Vision, Build Systems, Linux, Optimization, Signal Processing, Matrix Algebra, Algebra, Geometry, Template Metaprogramming, Unit Testing, Agile Workflow, Test-driven Development (TDD)

Computer Vision and Machine Learning Engineer

2018 - 2020
Cubist IT
  • Developed a truck trailer detection and tracker algorithm based on images and vehicle odometry data on CAN.
  • Built a multi-camera object detection and tracking system for a construction equipment company.
  • Trained, pruned, quantized and deployed DNN models for object detection on an embedded FPGA accelerated platform.
Technologies: C++, Artificial Intelligence (AI), Bayesian Statistics, Computer Vision, Linux, Neural Networks, Machine Learning, Python, Sensor Fusion, TensorFlow, 3D Graphics, Build Systems, Geometry, Template Metaprogramming, Unit Testing, Test-driven Development (TDD)


2016 - 2017
  • Developed unsupervised recommendation systems in MATLAB.
  • Interfaced with MATLAB MEX C interface for different time-critical tasks. Optimized a sorting algorithm that was the bottleneck in the system. The solution used heap sort and multi-threading.
  • Batched large matrix operations that required more memory than available RAM.
Technologies: MATLAB, Python, C, Optimization, Machine Learning, Neural Networks, Unsupervised Learning, Matrix Algebra, Non-negative Matrix Factorization (NMF)

Automatic Trailer Tracking

I contributed to setting up a whole deep learning training pipeline to detect truck trailers using TensorFlow. On top of that, I developed a recursive Bayesian estimation algorithm for inferring the trailer's angle over time using various sensors such as the camera and vehicle odometry on the CAN bus. Finally, I used tools to deploy the embedded system, including the FPGA acceleration of the neural network.

Open-source C++ Library
I developed a small library for handling sets of enumerations or elements of any fixed size collection in a type-safe way. The library is fully integrated with both the build2 and CMake build systems.
2011 - 2018

Master's Degree in Applied Physics | Electrical Engineering | Computer Vision

Linköping University - Linköping, Sweden






Test-driven Development (TDD), Unit Testing, Template Metaprogramming, Agile Workflow


C++, Python, C




Computer Vision, Machine Learning, Optimization, Artificial Intelligence (AI), Neural Networks, Signal Processing, Bayesian Statistics, Sensor Fusion, Matrix Algebra, Geometry, Deep Learning, 3D Graphics, Build Systems, Unsupervised Learning, Non-negative Matrix Factorization (NMF), Algebra, Epipolar Geometry

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.


Share your needs

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

Choose your talent

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

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