<aside> 🪐
Hi, I’m Ana, an Aerospace Engineering MSc student, passionate about artificial intelligence, deep tech and autonomous space systems.
This portfolio highlights my main interests and experiences in the fields of ML, intelligent control systems and distributed multi-agent applications.
</aside>
I am 24, passionate about AI, computer vision, and aerospace control. I dream of connecting cutting-edge innovation with real-world challenges, and of collaborating in pushing the frontier in AI, automation and spacetech. My bet is on using interconnected satellite constellations to improve life on Earth, respond faster to disasters and get to know our planet better. 🚀
I am a MSc student in Aerospace Engineering at TU Delft, with a focus on machine learning, computer vision and autonomous multi-agent systems. I am currently pursuing a Visiting Research Period at the ESA Φ-lab ****as part of my master’s thesis on “Multi-Agent Coordination for Task Orchestration on Satellite Constellations” where I am developing a framework for coordinating wildfire monitoring observations from different satellites and fusing them onboard an AI orchestrator satellite in order to aid decision making on ground.
Born in Brașov 🇷🇴, I developed my technical foundation in Delft 🇳🇱, expanded my global perspective in Singapore 🇸🇬, and now continue building in Paris 🇫🇷 and researching in Rome 🇮🇹.
Description: My MSc thesis investigates the use of VLM-based reasoning models for autonomous multi-satellite coordination in disaster management by combining onboard wildfire detection with satellite situational awareness for collaborative Earth observation tasking.
Code: coming soon
Description: 3D interactive explorer where users evaluate data center placement, including on-orbit. Powered by AI-generated feasibility reports, blockchain-verified portfolio records, and real sustainability data, built in 48 hours at HackEurope Dublin 2026.
Code: http://github.com/viftode4/hackaton-dublin

Description: Building an agent to handle conversations and autonomously schedule patients for consults
Code: not accessible
Description: Designed a Fleet of Autonomous Drones to Measure Cargo in Warehouses
My work on Operations: https://photos.app.goo.gl/qnseQJgs4ids8m8d8

Description: Build a 3D reconstruction pipeline to take in a few poses, neurally extrapolate the object and reproject in 3D through structure from motion
Code: https://github.com/anapelin/3DHousePlan

Description: Analysis of late-night dining & transportation options in relation to Singapore nightlife hotspots

Code: https://github.com/anapelin/happysip_happyslurp_happyshuttle
Description: Built and deployed detection, stereo and depth DNN models. Improved the speed of preprocessing by 50%.
Code: not accessible
Description: Integrated a drone flight simulator with perception and control products. Performed SITL and HITL.
Code: not accessible
Description: Won 4th longest flight place in a drone obstacle avoidance competition. Designed object avoidance controllers with depth estimation and traditional computer vision.
Description: Designed a mission to Venus to scan the atmosphere of the planet + full system design.
Description: Designed operational databases and developed telemetry commands for the comms subsystem of DA Vinci’s Educational satellite.
https://davincisatellite.nl/nl/
Description: Designed a sensor bias estimation algorithm and noise filter for spacecraft 3D attitude control

Description: Built an algorithm to optimally distribute a fleet of drones and trucks to deliver packages in Delft

Description: Designed a compressed gas regulator to handle the pressurant lines of the firebolt engine.
