ANKUR KOHLI
Software Engineer
I’m a Software Engineer focused on building scalable,
intelligent, and high performance systems that blend clean architecture with real world
functionality. With expertise in Python, C/C++, Java, JavaScript, Dart (Flutter), and modern
development practices, I design and deploy cross platform applications that run seamlessly
across mobile, web, and desktop.
I have extensive experience designing and deploying realtime systems with Dockerized ROS2, CI/CD
pipelines, and GitLab automation, ensuring reliability, modularity, and hardwareagnostic
scalability. I am skilled in AI, Machine Learning, and Computer Vision using PyTorch,
TensorFlow, and OpenCV, leveraging these technologies to enhance automation and data driven
decision making. I specialize in API driven clientserver development, realtime systems, and ROS2
based backends, enabling efficient, low latency communication between components. From intuitive
Flutter UIs to robust backend integrations, I deliver software that is reliable, responsive, and
built for long term scalability.
Passionate about writing clean, maintainable code, architecting efficient backend systems, and
creating intuitive user experiences. I value collaboration, continuous improvement, and staying
up
to date with emerging technologies always eager to learn more, expand my skills, and take on new
challenges to deliver smarter, more innovative software solutions.
Career Identity
Aspiring Software Engineer with a strong foundation in Python, C/C++, Dart, Flutter, API driven client server development, Java, Bash, Powershell, & advanced technologies.
Professional Experience
Software Engineer April 2025 - October 2025
NTT Data Italia Italy
- Developed cross platform Flutter frontends with a UI/UX focus and realtime, multithreaded ROS2 backends, ensuring secure, low latency communication for intelligent robotic systems.
- Built and deployed autonomous SLAM based mapping and navigation pipelines with Dockerized ROS2, CI/CD workflows, and real robot deployment, enabling hardware agnostic scalability and faster operations.
- Conducted software testing, validation, and debugging via GitLab CI/CD and GitLab Issues, improving system reliability and robustness.
- Authored technical documentation and architecture diagrams, enhancing team collaboration, maintainability, and adherence to software engineering best practices.
Thesis Project June 2023 - February 2024
NTT Data Italia Italy
- Developed realtime, multithreaded arch to ensure low latency performance and improved system efficiency by 30% for intelligent robotic systems.
- Prototyped SLAM-based mapping and navigation pipelines within Dockerized ROS2 environments, establishing a foundation for scalable CI/CD deployments.
- Tested, validated, and debugged software using GitLab CI/CD, increasing system reliability and uptime by 20%.
- Authored comprehensive technical documentation, improving team productivity and maintainability across projects.
Projects
Restaurant Website
A Restaurant website developed using HTML, CSS, and some minor functions in JavaScript, demonstrating front-end development skills and basic web design principles.
3D reconstruction
To developed immersive virtual reality experiences of robotics by capturing and converting real-world objects or environments into detailed 3D models.
Cervical Cancer Detection using CNN & VGG16 Module
This work is based on Cervical cancer detection using CNNs involves leveraging deep learning models like VGG16 to classify cervical cancer in medical images, such as Pap smear slides, due to its proven effectiveness in image recognition tasks.
Software Architecture for Mobile Robot Control
This project involves developing a software archi- tecture for controlling a mobile robot by applying graph-based routing and Dijkstra’s. It is based on using ROS, Gazebo, and RViz to control the robot. Also, it consists of writing ROS nodes in Python: a controller for the robot and a UI.
Reinforcement Learning using A2C Algorithm & PPO Algorithm for Lunar Lander
Implemented solutions for the Lunar Lander problem using Advantage Actor-Critic (A2C) and Proximal Policy Optimization (PPO) algorithms within the framework of Reinforcement Learning, focusing on optimizing policy by implementing policy iteration approaches and value functions for efficient and robust lander control.
