Igor Pejic, Developer in London, United Kingdom
Igor is available for hire
Hire Igor

Igor Pejic

Verified Expert  in Engineering

Software Developer

London, United Kingdom
Toptal Member Since
April 2, 2018

Igor is an experienced CTO with over seven years of development expertise. He has a master's degree in artificial intelligence and enjoys the utilization of machine learning in software development. Igor prefers to work on scalability and performance-intensive projects. He holds a work authorization for the UK as a Global Talent visa holder and in the EU as an EU citizen.


PHP, Mercurial, React, Python 3
Python 3, Apache Airflow, React, Django, SQLAlchemy, PostgreSQL, Google Cloud...
Plotly, Docker Swarm, Ansible, PostgreSQL, Redis, InfluxDB, Linux, Systemd...




Preferred Environment

Git, Shell, Vim Text Editor, Linux

The most amazing...

...program I've developed is a computer vision project to identify the same person in different parts of the store.

Work Experience

Solutions Engineer

2022 - 2023
  • Created data pipeline for ePD data business analytics by anonymization and aggregation.
  • Led and developed beta launch in the EMEA region for new business messaging features, including WhatsApp Business search, WhatsApp Flows, and recurring notifications on Instagram.
  • Managed book of business for Meta's partners, generating over $20 million in annual revenue.
Technologies: PHP, Mercurial, React, Python 3

Lead Software Engineer

2019 - 2022
  • Managed the product team of five people and reported to the board of directors. The product is a SaaS solution with machine learning capability to help teachers provide better service to pupils in primary schools.
  • Led development of the full-stack codebase using Kubernetes, Google Cloud, React, PostgreSQL, and Python.
  • Increased the nightly dump sync performance and reliability by optimizing the database schema and increasing visibility and error recovery.
Technologies: Python 3, Apache Airflow, React, Django, SQLAlchemy, PostgreSQL, Google Cloud, Kubernetes, Terraform, Pandas


2017 - 2019
  • Led a team of 10 developers with expertise in the front end, back end, DevOps, and R&D. Supported the team's growth, enhanced CI practices, promoted test-driven development, implemented agile methods, and met deadlines.
  • Migrated the application architecture to a container-based, self-healing, scalable solution to accommodate growing traffic and the need for stability.
  • Improved development speed and released useful features by introducing new appropriate open-source libraries and continuous integration best practices.
  • Constructed a monitoring solution with error prevention, real-time monitoring, and automatic alerting for a multi-location architecture, resulting in reduced downtime, enhanced stability, and optimized resource utilization.
Technologies: Plotly, Docker Swarm, Ansible, PostgreSQL, Redis, InfluxDB, Linux, Systemd, Keepalived, NGINX, Keras, TensorFlow, Lua, Docker, React, C++, Python

R&D Engineer

2016 - 2017
Monolith International B.V.
  • Led computer vision research utilizing machine learning and state of the art algorithms.
  • Developed a person re-identification solution able to detect the same person on multiple distant camera views.
  • Improved accuracy and capabilities of proprietary person behavior-tracking algorithms using object detection, segmentation, localization and classification.
  • Collected a training dataset and utilized it to build multiple custom neural networks which are still used in production.
Technologies: Amazon Web Services (AWS), NVIDIA CUDA, Microsoft Azure, Google Cloud, Keras, Dlib, TensorFlow, OpenCV, C++, Python

Software Engineer

2015 - 2016
Logit Ltd.
  • Developed a plugin payment gateway which connects to the payment system of the biggest telecom company in Croatia.
  • Created an application for tracking work hours of employees. The solution replaced an excel-based system saving time for data entry and generation of reports for the tax office.
  • Improved the ease of use of a web application for automatic generation of reports by redesigning it as a single-page application.
Technologies: JSON, ReportLab, jQuery, XML, OAuth, REST, Django CMS, Flask, MySQL, AngularJS, Python

Software Engineer

2014 - 2016
  • Developed a search plugin for a CMS website using ElasticSearch.
  • Created a solution for worker and spaces occupancy timetables.
Technologies: Django, PostgreSQL, Django CMS, Python, Redis, Elasticsearch

Software Engineer

2014 - 2015
  • Contributed to a web application solution used by the university to input student grades and sync them to the central database for educational data. The solution is still in use today for 10,000 students on a yearly basis.
  • Implemented single sign-on authentication.
  • Managed a system able to sync to and from central database.
  • Implemented reporting to pdf using ReportLab.
  • Deployed, monitored, and managed on linux.
Technologies: ReportLab, Sphinx Search Engine, Shell, jQuery, XML, Flask, NGINX, Apache, Django, PostgreSQL, Python

RESTful API for Android Application

RESTful API with token based authentication, restful principles connected to a PostgreSQL database.

How I Stay More Productive by Using the Mouse Less

An article in which I described my development environment and set up aimed towards using Mouse Less to be more productive while programming. Contains tips on how to set up your Linux environment to do so.

Stackoverflow Profile (Top 11% of all Users)

I enjoy contributing to the tech community by sharing expertise in areas such as C, C++, JavaScript, Python, Django, and software engineering practices. On my profile, I actively engage by asking questions, providing answers, and participating in discussions.

Scientific Publication - Monte Carlo Tree Search on Perfect Rectangle Packing Problem Instances

We explored the possibilities of Monte Carlo tree search (MCTS), which showed immense success in games like Go and Chess in solving the perfect rectangle packing problem.

We performed experiments with two differently generated problem sets of 1,000 instances each and explored six different rollout numbers and two different action-selection strategies for MCTS. Then, we compared the algorithm's performance to an exact depth-first algorithm equipped with efficient pruning techniques. By rating the number of solutions found against the total number of tiles placed, we defined a 'computationally economic tradeoff'. Different rollout numbers and strategies lead to different results, both within and between the two problem sets. We discussed these results in the context of other heuristic algorithms on this problem and closely related areas.
2018 - 2020

Master's Degree in Artificial Intelligence / Software Engineering

University of Amsterdam - Amsterdam, Netherlands

2013 - 2016

Bachelor's Degree in Computer Science

Technical Faculty of Rijeka - Rijeka, Croatia


Django ORM, Keras, jQuery, ReportLab, OpenCV, Dlib, React, Jenkins Pipeline, Graphene-Django, TensorFlow, SQLAlchemy, Pandas, PyTorch, Scikit-learn


Vim Text Editor, Shell, Docker Swarm, Docker Compose, Ansible, NGINX, Git, Systemd, Plotly, AppOptics, Amazon Virtual Private Cloud (VPC), AWS Batch, Amazon Elastic Container Service (Amazon ECS), Amazon CloudWatch, Amazon EBS, Amazon Elastic Container Registry (ECR), Jenkins, Graphene, Sentry, Docker Hub, Keepalived, Flow, Apache, Makefile, Mercurial, Apache Airflow, Terraform


Django REST Framework, Flask, Redux, Django, OAuth 2, AngularJS, Angular


JavaScript, Python 2, Python, Python 3, GraphQL, XML, Haskell, Erlang, Lua, Go, C, C++, HTML, PHP




PostgreSQL, JSON, MySQL, Sphinx Search Engine, Google Cloud, InfluxDB, MongoDB, Memcached, Amazon S3 (AWS S3), Elasticsearch, Redis


Django CMS, NVIDIA CUDA, Amazon Web Services (AWS), Linux, Debian, Docker, Apache2, Kubernetes


Cython, Serverless, Microsoft Azure, SOAP, OAuth, Relational Database Services (RDS), Productivity, Machine Learning, Natural Language Processing (NLP), Scientific Computing

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