PyData Eindhoven 2025
2025-12-09
Eindhoven, Netherlands
[PyData Eindhoven 2019, PyData Eindhoven 2020, PyData Eindhoven 2021, PyData Eindhoven 2022, PyData Eindhoven 2023, PyData Eindhoven 2024, PyData Eindhoven 2025]
| title | author | length |
|---|---|---|
| Official Aftermovie | 01m | |
Quick montage of crowd and stage highlights that confirms scope rather than substance: footage shows an estimate of a few hundred attendees, three parallel… | ||
| Opening | 06m | |
Welcome and logistics for PyData Eindhoven 2025 name the venue as High Tech Campus, note sponsorship and stewardship by NumFOCUS for projects like NumPy and… | ||
| Closing | 08m | |
PyData Eindhoven 2025 closes with concrete numbers and outcomes rather than promises: 27 speakers 2 keynotes 250 attendees, 30 volunteers unpaid, and €75,000… | ||
| organizing team aftermovie | Dmitry Bagaev | 03m |
An aftermovie created by a core team member compiles event footage into a short music-driven highlight reel, opening with visible timestamps at 00:00 and 00:10… | ||
| Planning Hockey Careers With Python | Jaroslav Bezdek | 31m |
A deployed career-planning feature uses historical player paths from ~25,000 profiles, filters for same start age and league, builds nodes as age+league pairs… | ||
| From Experiment to Enterprise: Architecting AI for Stability and Scale | Roland Kunz | 29m |
Claims concrete patterns: an Dell AI Factory that stitches silicon, models, and software into a repeatable pipeline; a tested hub at dell.huggingface.co with… | ||
| Optimizing fantasy basketball decisions with Python | Pawel Kapuscinski | 29m |
Shows how PuLP and linear & integer programming encode roster rules and position eligibility, with three concrete Python examples: a start/sit optimization, a… | ||
| From Data Lake Entanglement to Data Mesh Decoupling: Scaling a Self-Service Data Platform | Geert Jongen | 30m |
Claims conversion from a monolithic EDW to a domain-oriented data mesh using PySpark Delta Lake DQX and Unity Catalog rests on measurable changes: about… | ||
| Extending SQL Databases with Python | Florents Tselai | 31m |
Shows how Python actually runs inside three databases—SQLite, DuckDB, PostgreSQL—by registering functions with sqlite3.create_function, building custom… | ||
| Identifying playstyles in football through spatial networks | Annemarijn Blom | 32m |
Claims an automated pipeline converts tracking and event data (25 Hz) into spatial networks, computes seven metrics like convex hull, opponent/covering degree… | ||
| Finding trash in waste | Tom Koopen | 31m |
Claim: a camera-based top‑view system reduces subjective rejection by automatically flagging contaminants; currently 8.5% of truck loads is rejected. Pipeline… | ||
| From €1M License to In-House Success: How We Built a Real-Time Recommendation System and Saved Millions Doing It | Ali Kohan | 30m |
Company replaces a €1M license by building a Kotlin real‑time recommender, deployed to 5% traffic then scaled to millions of requests per second, with an… | ||
| Developing a Nation-Wide Padel Rating System: A Data-Driven Approach | Max Brouwer | 30m |
A new padel rating replaces tennis tooling with a data-first system trained on 300,000 matches, compares Elo variants and a game‑index model, and uses… | ||
| Scaling Retail Planning at IKEA: Orchestrating Sales, Fulfillment and Capacity Assessment with Metaflow | Yannick Mariman, Harlley de Lima | 31m |
IKEA orchestrates sales, fulfillment and capacity forecasting into a single pipeline using Metaflow + Argo Workflows to run parallel jobs for 30+ countries… | ||
| Beyond One Model: Scaling, Orchestrating & Monitoring | Azucena Morales, Vi Chu | 31m |
Shows a deployed anomaly-detection fleet that moves from a single model to per-shuttle models, claims scaling to hundreds (and a path to thousands) by running… | ||
| Scaling Python to thousands of nodes with Ray | Rob de Wit-Liezenga | 31m |
Claims ability to run Python workloads across thousands of nodes using Ray, demonstrating actual code changes and autoscaling from a laptop to a cluster that… | ||
| CompactifAI: Quantum-Inspired AI Model Compression | Jon Leiñena Otamendi | 30m |
Claims a pipeline that uses tensor networks to profile weight correlations, surgically prune redundant parameters, then run a tiny healing step (one to two… | ||
| Ingestify: Rethinking Ingestion for Complex Data | Koen Vossen | 27m |
Ingestify presents a concrete system: a Metadata-First Ingestion Layer that stores raw data as-is (S3 or local objects, Postgres/SQLAlchemy metastore) and… | ||
| Federated Data, Centralized Action: A Governance Model Powered by Lakewatch and Databricks System Tables | Cristiano Cortez Da Rocha, Frank Chidi Mbonu | 30m |
Claim: a three‑pillar pipeline turns Databricks System Tables into continuous, actionable governance by extracting usage, aggregating with Lakewatch and… | ||
| Efficient Time-Series Forecasting with Thousands of Local Models on Databricks | Daria Mustafina | 28m |
Shows scaling local models on Databricks using Pandas API on Spark, training thousands of per‑group models in parallel; backtesting runs 20,000 parallel folds… | ||
| xReceiver: a GNN approach to the evaluation of the decision-making process of passing options in football | Gabriel Masella | 25m |
xReceiver is a Message‑Passing GNN that models teammates as nodes and potential passes as weighted edges (distance, angle, opponent pressure), trained on… | ||
| Hier is Vincent. Eén en al oor. [Closing Key note] | Rob Mulder | 34m |
Claims a Digital Twin built from 800+ Van Gogh letters, uses LLMs plus topic‑context emotion mapping to produce time‑sensitive responses, and links answers to… | ||
| Building, Deploying and Managing AI Agents at Scale | Sander van Donkelaar | 31m |
A pragmatic engineering blueprint for building, deploying, and managing agentic AI on Google Cloud, focusing on scalable, secure, and cost-controlled… | ||
| The Science of Showing Up [Opening Key note] | Maurits Hendriks | 53m |
Key claims and concrete examples decide: a 144‑medal track record and a rule that Olympics are built every day not every four years; measured wins from an… | ||
| Football is complex, but your code doesn’t have to be — meet DataBallPy and a practical deep dive into pressing | Alexander Oonk, Tygo Nikamp | 31m |
DataBallPy is an open‑source Python package that actually gives a usable pipeline: a synchronization algorithm reduces timestamp error from ~4s to within one… | ||
| nothing matches | ||