PyData Fest Amsterdam 2020
2020-06-15..2020-06-20
-- online --
who it's for
data scientists, python developers, machine learning engineers, data engineers, MLOps engineers, data governance leads
topics
voice driven coding with Talon; python performance profiling with scalene; list comprehension micro optimizations; pandas performance improvements; Python PEG parser internals; serving models with FastAPI; federated learning with secure aggregation; time series forecasting with sktime; data validation with Great Expectations; dual online offline feature store; population shift monitoring with popmon; Spark CSV parsing pitfalls
| title | author | length |
|---|---|---|
| Accessible Means Hackable | Naomi Saphra | 47m |
A programmer loses the ability to type in 2015 and learns to dictate code with Talon, customizing macros, a phonetic alphabet, an indexed clipboard and repeat… | ||
| Interview | Guido van Rossum | 35m |
History and governance notes stop being vague and get concrete: explanation of why Python 3 adoption lagged — lots of throwaway code and non-programmer users —… | ||
| Keep the Data Flowing | Madelon Hulsebos, Tomislav Suhina, Joost Boonzajer Flaes | 01h 34m |
Workshop demonstrates deploying Great Expectations to automate data checks: init a data context, run the profiler, generate data docs and an expectation suite… | ||
| Privacy by Design Machine Learning | Romeo Kienzler | 44m |
Demonstrates a working pipeline that moves training to devices using federated learning, runs update aggregation in Python notebooks with an open‑source… | ||
| Introduction to Machine Learning with Time Series | Markus Löning | 01h 27m |
A practical primer shows a new Python toolbox sktime that actually implements reduction-to-regression, ensembling, pipelining, detrending and temporal CV… | ||
| Furious & Fast Python 7: Writing Fast Python Code | James Powell | 01h 12m |
Claims a concrete M-I-S-S-I framework (Measurement, Intuition, Structure, Specialization) and shows tools and data: benchmarks with scalene and pyspy reveal… | ||
| Human-like Visual Search Application with Small Data | Alyona Galyeva | 01h 29m |
Builds a Siamese Mask R-CNN for one‑shot instance segmentation, trained from COCO with 80 categories, uses pre‑trained checkpoints (large config ~1 week on 4… | ||
| High Performance Data Loss | Ondrej Kokes | 29m |
Concrete failures with Apache Spark parsing CSVs produce silently wrong results: invalid dates overflow to other calendar days, type mismatches nullify entire… | ||
| Serving ML Easily With FastAPI | Sebastián Ramírez | 45m |
FastAPI shows how to turn a working ML model (spaCy entity extractor in examples) into a production JSON API with minimal code, delivering automatic OpenAPI… | ||
| Supporting Business Ethics and Governance with Data | Max Gendler | 31m |
Describes a concrete system called Project Feels that models reader emotions from off‑site survey labels and attaches emotion tags to every article, then uses… | ||
| Popmon: Population Shift Monitoring Made Easy | Tomas Sostak | 32m |
A lightweight Python pipeline called popmon monitors covariate shift by histogramming features, applying tests like chi-square and Pearson/KS, and producing… | ||
| A feature Store for Machine Learning | Jim Dowling | 32m |
Feature engineering is productized: Hopsworks.AI provides a versioned DataFrame API and Apache Spark jobs that materialize feature groups into a dual… | ||
| Open-source in the eyes of a Minority Contributor | Cheuk Ting Ho | 27m |
Covers concrete wins and gaps: built a Python client for TerminusDB and runs frequent Twitch tutorials, started contributing at a Bloomberg sprint where… | ||
| Designing for Impact in Civic Data Science | Andrew Therriault | 35m |
Open data portals publish raw feeds—Analyze Boston exposes 160 datasets updating daily via CSV and an API with scripts that push every four hours—yet most… | ||
| Python: From Source to Execution | Chiel Peters | 27m |
Starts from a tiny snippet and traces concrete steps: the tokenizer produces tokens, the parser builds an AST, the compiler emits a code object with… | ||
| Making Pandas Fly | Ian Ozsvald | 29m |
Shows concrete wins: converting repeated strings to category dtype halves memory and cuts a value_counts from ~400ms to ~23ms, switching float64 to float32… | ||
| Exploring Railway Oriented Programming in Python | Bertjan Broeksema, Huib Keemink | 29m |
Builds a Railway Oriented Programming pipeline from scratch in Python and demonstrates the concrete code: a Success/Failure two-track wrapper, bind and map… | ||
| Making an Impact with Duct Tape | Thijs Suijten | 26m |
A hacker collective builds low‑cost field systems that actually work: an autonomous hobby drone for ~€100–200 replaces a €50k unit; a grid of sensors detect… | ||
| Watching Millions of Trees with Python: For Impact and Profit | Roelof Pieters | 36m |
Claim: machine learning on high‑resolution satellite imagery monitors millions of trees worldwide and surfaces actionable alerts. Evidence: 100,000 labeled… | ||
| nothing matches | ||