Elian Angius
Elian Angius

Elian Angius

Applied Data Science & Machine Learning Lead

Get in Touch View Projects

About

I work on creative & challenging data problems without clear answers — usually somewhere between maps, language & real-world behavior. I’m especially interested in the exploratory side of the work: shaping questions, experimenting quickly & engineering ideas into real solutions.

Skills

Topics I specialize in & enjoy working on:

Data Science & ML

Predictive Modeling Forecasting Deep Learning Agentic Systems Vector Embeddings Information Retrieval Model Explainability

Location Intelligence

Geospatial Mapping Kernel Density Heatmaps Search & Routing Demographic Modeling Knowledge Graphs

Applied Research

Exploratory Analysis Rapid Prototyping System Design Simulation A/B Testing

Leadership

Requirements Specs Agile Roadmapping Cross-Functional Collab Product Demos Team Mentoring

Projects

A selection of my works, experiments & prototypes:

Occupancy Prediction

Classifies daily booked, offline & vacant calendars of scraped Airbnbs — enabling accurate revenue projections.

  • Time-Series
  • Hierarchical Classification

Vacation Rental Valuator

Appraises properties relative to their markets — trained on-demand allowing for what-if competitive analysis.

  • Simulation
  • Explainability
  • Ranking

Safer Routing

Navigating pedestrians across a city through the safest path — not just the shortest.

  • REST
  • Geospatial
  • Optimization

Neighbourhood Vibes

Measure the local sentiment & interests from streaming social media, news & events.

  • Simulation
  • NLP
  • LLM
  • Geospatial

Probabilistic Geocoder

Resolves noisy, ambiguous & privacy-protected location references from web data into a best-effort polygon & confidence-weight.

  • NLP
  • Geospatial
  • Probabilistic Modeling

ML Debuggability

Instruments tree-based models with learn-time metadata & run-time signals to diagnose why any prediction went wrong.

  • MLOps
  • Explainability
  • Tree Models