Elif Pulukcu
Jul 2026
Kayaking & Canoeing
One of the best things I've added to my life in 2026: kayaking and canoeing, my new hobbies. Kanonbådsvej is a great spot for it, but Mølleå is still my favourite.

I'm an MSc student in Autonomous Systems at the Technical University of Denmark, with a background in Computer Engineering.
I am interested in intelligent systems that connect perception, decision-making, and reliable software engineering. What fascinates me most is how autonomous systems understand the world around them and turn data into meaningful actions.
My work often sits between algorithms and implementation: building systems that are not only technically interesting, but also testable, deployable, and maintainable in real-world environments.

Studies and projects, events and volunteering, travel, hobbies. In short: life.
› Current weighted grade average: 10.11 out of 12
Courses I’ve Enjoyed Most So Far
Teaching Assistant Experience: stay tuned for the details!
Activities and Societies: DTU Salsa
› Graduated 5th in the Faculty of Engineering
Thesis: Machine Learning-Based Web Application Firewall
Minor Degree: Electrical, Electronics and Communications Engineering
Activities and Societies: Board Member of Computer Engineering Club, Member of IEEE Robotics and Automation Society
› Computer Vision System for Product Attribute Recognition and Outfit Recommendation
› Natural-Language Database Query System (text-to-SQL RAG)
Letting anyone query the database in plain English, no SQL required.
› Retail Data Analytics & Reporting
Turning raw retail data into reports and dashboards business teams actually use.
Worked across machine learning, industrial IoT and data engineering for predictive maintenance and production traceability, moving real-time factory data all the way from shop-floor devices to cloud deployment on AWS and Docker.
› Industrial IoT Predictive Maintenance System
Catching machine failures before they happen, from raw sensor streams to live alerts.
› Production Traceability for Washing Machine Manufacturing
Following every washing machine down the line to trace exactly where things go wrong.
Worked mostly with automotive data, building computer vision and machine learning solutions end to end and training models on AWS.
› Automotive Dashboard Warning Light Recognition
Teaching a model to read the cryptic symbols on a car's dashboard.
› Automated Labor Planning
Helping service centers staff the right number of technicians on the right days.
› TV Ratings Forecasting
Predicting who’s tuning in, episode by episode.
A few things I’ve built. More on my GitHub.
A loosely-coupled GNSS/INS navigation system that fuses satellite positioning with IMU data through a 15-state error-state Extended Kalman Filter. Features Single Point Positioning, Allan-deviation IMU characterisation and GNSS outage simulation, tested on real UGV datasets.
A real-time system that tracks multiple vessels in a harbour by fusing mm-wave radar, stereo camera, AIS and GNSS through an Extended Kalman Filter, with Hungarian data association and an M-of-N track lifecycle. Validated on simulations and real Copenhagen harbour recordings.
A semantic segmentation model (nnU-Net) that classifies urban drone imagery into obstacles, water and landing zones for safer autonomous landings, wrapped in a full MLOps pipeline with DVC, Docker, BentoML and CI/CD on Google Cloud.
Graph Neural Networks for molecular property prediction on the QM9 dataset, comparing fully supervised training with semi-supervised approaches (Mean Teacher and N-CPS) to learn from unlabelled molecules.
A GeoCLIP-inspired retrieval system that localizes images by embedding ground photos, satellite imagery, text and GPS into a shared contrastive space. An ablation across nine modality combinations cuts median error from ~500 km to ~124 km, extended with cross-view retrieval and University-1652 transfer.
A multi-modal route planner for Istanbul's railway network (metro, tram, funicular, cable car). Computes efficient itineraries with an A* search over a GeoJSON station graph, minimising transfers, served through a Flask API with an interactive Leaflet map.