Interview with Milan Janosov
summary
Claims that geospatial work fails because people treat GIS and ML as separate worlds are front and center, arguing the real task is connecting GIS to Python/ML. Practical rollout stalls on raw inputs: randomly chosen geospatial datasets are “crap” and typically require many hours of cleaning before pipelines accept them. Backgrounds in physics or biophysics transfer as a quantitative, problem‑solving mindset. Emerging AI is framed as a prompt-driven way to auto-generate Python snippets and cut debugging time.
AI-written summary · about
chapters (5)
- 0:00Start
- 0:07Can you tell me what are some of the misconceptions and misunderstandings that people have of Geospatial Data Science?
- 0:36What are some of the main challenges that you face when using these models in businesses and then introducing them and having a practical use for them?
- 1:18How do you think your background in physics and biophysics helped you with getting into and understanding data science?
- 2:08What excites you the most what prospects do you see as really exciting when it comes to both AI and Data Science?
related talks
50mGeospatial Analysis
33mBuilding Data Science Tools for Sustainable Transformation
30mLocation Intelligence and Spatial Data Science: from geographic coordinates to business insights
01h 36mFrom Maps to Models: An End-to-End Journey in Geospatial Machine Learning with Python
32mThinking Like a Data Scientist
46mData Discussion - AI for Good
05mMaking data science useful.
34mForget about AI and do mathematical modeling instead!