1 år i yrketEgenföretagare

Hugo Asztely

Data Engineer

Göteborg

Hugo Asztély is a data engineering consultant with a background in machine learning, AI and fullstack development. He works across the whole chain, from collecting and cleaning data to building models, AI services and the applications people use them in. He cares about data that can be trusted and solutions that hold up in production, and has experience from regulated environments where security and compliance matter.

At Elevyo Health he develops a clinical platform for healthcare providers, combining fullstack development with AI agents, similarity analysis and monitoring of patient data. In his master's thesis at Sweco he showed that machine learning could improve the estimates the Swedish Transport Administration relied on. He works well in small teams, close to both business and technical colleagues.

Visa mer

Kontakta

HTML/CSS/JS
Matplotlib
Python
MySQL 5.7
PostgreSQL
Docker
Redis
Tailwind
MATLAB/Simulink
NumPy
Git bash
Communication
Data Modelling
Data Quality
Algorithms and Data Structures
Machine Learning & AI
AI Agent Development

Kurser & certifieringar

FME Form BasicFME · 2025
Integrate Data with the FME PlatformFME · 2025

Uppdrag

Elevyo Healthfinder
jan. 2026 – nu
Developer Göteborg
Elevyo Health is a medtech startup building Elevyo Healthfinder, a B2B platform for health tracking used by healthcare providers. Hugo is a fullstack developer in a small team with the two founders and one other developer, and takes on different roles depending on what the company needs.
Fullstack: Develops the multi-tenant Django backend and Nuxt/Vue and Htmx frontends, with GDPR and EU MDR requirements in mind.
AI agents: Develops agents that summarise patient reports, and a pipeline that processes academic PDFs with pdfplumber and GROBID into pgvector embeddings for similarity search.
Similarity and clustering: Develops microservices that compare patient scores with patient and research data to show similarities, and works on clustering and trend analysis of patient groups.
Trend monitoring: Develops dashboard content that monitors an organisation's patients over time and highlights those with the worst trends or values.
Logging, testing and automation: Works on audit logging, automated testing and automated workflows for the platform's users, with scheduled jobs in Celery and CI/CD in GitHub Actions.
Visa merVisa mindre
- Python
- PostgreSQL
- Django REST framework
- Celery
- Docker
- Git / GitHub / GitHub Actions CI/CD
- RAG (LLM)
- AI Agent Development
- Nuxt
- Vue
- Htmx
- Tailwind
Sweco Digital Services
jan. 2025 – juni 2025
Master's thesis Stockholm
The thesis examined whether machine learning could predict clearance times for traffic accidents in Stockholm more accurately than the manual estimates used by the Swedish Transport Administration.
Data pipeline: Built an FME and QGIS workflow combining accident data from NTS with weather data from SMHI.
Data quality: Identified and excluded estimates that were system defaults rather than real assessments, and filtered outliers using the interquartile range.
Feature engineering: Extracted keywords from sparse free-text fields and one-hot encoded them as features.
Modelling: Benchmarked several models, where Random Forest and XGBoost performed best, evaluated with MSE and MAE against the manual estimates.
Result: Showed that machine learning improves on the manual estimates. Published in DiVA.
Visa merVisa mindre
- Python (NumPy; Pandas; scikit‑learn)
- FME (Feature Manipulation Engine)
- QGIS
- Matplotlib
- Seaborn
- XGBoost & Random Forest

Utbildning

Civil Engineer
2020
Uppsala University

Kompetens · nivå och år

Tekniker

HTML/CSS/JS
Matplotlib
Python
User experience
AI
C#
Celery
CSS
Data Visualization
Django
Django REST framework
JavaScript
Seaborn
SQL
C++
Java
Stripe
TypeScript
Unity
Language / Translation

Plattformar

MySQL 5.7
PostgreSQL

Produkter

Docker
Redis
Tailwind
Keras
QGIS

Verktyg

MATLAB/Simulink
NumPy
Git bash
Microservices architecture
PostgresSQL

Metoder och processer

Communication
Data Modelling
Data Quality
Machine Learning
Agile Methods
Knowledge sharing

Övrigt

Algorithms and Data Structures
Machine Learning & AI
AI Agent Development
AI Design
AI Development Supervisor and Architect
Anomaly Detection
Automated workflows
Compliance and Security
Coolify
ETL-pipelines
FME (Feature Manipulation Engine)
Geographical Data
Git / GitHub / GitHub Actions CI/CD
GROBID
pdfplumber
PostgreSQL (pgvector)
Probability and Statistics
RAG- & LLM integration
Sci-kit learn
Traffic Data Analysis
Vue och Nuxt
XGBoost & Random Forest
API- & webhook integration
Bash / Unix / Linux
Django Allauth
Hetzner
Htmx
HuggingFace Transformers
LLM Integration
MCP (Model Context Protocol)
Ollama
PyTorch/TensorFlow
REST API Development

Den som känner Hugo

Sarvi Glemfors

Sarvi är Hugos kontaktperson på Kvadrat.

Mejla Sarvi