PyConZA 2019
| title | author | length |
|---|---|---|
| Principals of System Integration | Matthew French | 40m |
Real systems often break in predictable ways and the talk gives concrete fixes: use a queue as store-and-forward (demo swaps REST for RabbitMQ so outages don’t… | ||
| Anomaly Detection using Autoencoders | Naledi Modise, Angela Lai King | 28m |
Autoencoder pipeline uses Spark to H2O conversion, trains an h2o deeplearning estimator autoencoder on the open credit‑card dataset (0.17% fraud), saves the… | ||
| Production-ready Flask & Django apps on Kubernetes | Jamie Hewland | 33m |
Explains exactly how to run Flask/Django WSGI apps on Kubernetes: use a sidecar nginx proxy (or async Gunicorn/uvicorn workers or a service mesh) to protect… | ||
| Making Sense of Python Virtual Environment Tools and Workflows | Daniel Holmes | 37m |
Compares specific tools and behaviours: venv links vs virtualenv copies (venv is stdlib since Python 3.3), virtualenvwrapper stores envs centrally… | ||
| Sentry as a way not to be afraid | Irina Grechikhina | 38m |
Shows how Sentry captures real exceptions from a tiny Flask+SQLite app, reveals that a TypeError came from lists-of-lists via expanded stack-frame local… | ||
| Detecting Propaganda in Fake News using Natural Language Processing | Aroma Rodrigues | 38m |
A prototype extracts top phrases with RAKE keyword extraction, queries Google News API for matching articles, and measures overlap with spaCy similarity; it… | ||
| Using Python to visualize and analyze ion beam experiment data at a particle accelerator facility | Caswell Pieters | 37m |
Pipeline reads ROOT files via PyRoot into Python notebooks, uses Plotly graph objects for interactive scatter, contour and surface plots, and sends live… | ||
| Machine Learning Engineering principles with Python and MLFlow | Natu Lauchande | 43m |
Recreates a Michelangelo/Bighead-style pipeline using Python and MLflow to handle training, deployment, serving and champion/challenger tests, packaging models… | ||
| Managers, Diplomats & Gaskets: On Being the Seal That Holds Teams Together | Petrus Theron | 46m |
Argues that as domains deepen, the edges between islands of knowledge gain outsized value and the commercially useful role is a generalist who acts like a… | ||
| Automating infrastructure at SA Home Loans with Python (and friends) | Kim van Wyk | 34m |
SA Home Loans automates its 50-plus server production image into ~18 isolated labs (about 900 VMs total), replacing week‑long manual builds with Packer… | ||
| No more struggles with Apache Spark (PySpark) workloads in production | Chetan Khatri | 36m |
Covers concrete fixes: use PyArrow to avoid Python/Numpy/Pandas serialization bottlenecks, prefer JDBC parallel reads with numeric lower/upper bounds and… | ||
| Radio Astronomy Reduction Pipelines Using Python and Various Container Technologies | Athanaseus Ramaila | 44m |
Framework called Stimela packages radio‑astronomy tools as container images (built from PyPI and Debian/PPA artifacts) and runs them via… | ||
| Why You Should Care About Types | Luka Sterbic | 57m |
Argues that adding type annotations makes large Python codebases easier to read, speeds onboarding and prevents guesswork; shows incremental workflow that… | ||
| Getting Machine Learning Models Ready For Production Using Python | Adit Mehta | 35m |
Describes an eight-week production cycle that moves notebooks to a Docker-served model by splitting work into sprints: literature review, EDA (pandas… | ||
| Using QPython in smartphones as mobile computational laboratory anywhere, anytime and anyhow | Godfrey Akpojotor | 33m |
Demonstrates that QPython on Android phones runs real numerical lab work: a mass-spring model of a chick heart, a quadratic discriminant routine handling… | ||
| Deep Neural Networks for Video Applications | Alex Conway | 59m |
Shows concrete pipelines and results: uses pretrained convolutional nets (ImageNet->ResNet), transfer learning in Keras/TensorFlow to get 96.3% on ~3.5k… | ||
| Teaching kids to code 2.0 | Whitney Tennant, David Campey | 35m |
Summary of concrete outcomes and tools rather than promises: a volunteer-run CoderDojo network is turned into a local nonprofit CoderLevelUp that runs free… | ||
| Apache Airflow @ Umuzi.org | Sheena O'Connell | 36m |
Umuzi uses Apache Airflow to replace brittle cron and ad hoc scripts that handled thousands of applicant records, moving to dynamic DAG generation… | ||
| Working with data in Python for the Aspiring Data Scientists | Jean-Pierre Voogt | 35m |
Shows practical Python data work using Jupyter Notebooks with pandas DataFrame plus matplotlib/Seaborn, loading a zipped CSV of 2.7 million rows (~300–400MB)… | ||
| A gentle introduction on Building an intelligent Chatbot in Python | Paulus M Shituna | 35m |
Builds a simple rule-based chatbot using AIML and an AIML editor plugged into a Python workflow, showing a live demo with pattern/template pairs, random… | ||
| Programming for accessibility | Rory Preddy | 48m |
Argument: build accessibility from design not bolt-on, with concrete tooling and CI. Uses Accessibility Insights + its FastPass checks inside Chrome/Edge, runs… | ||
| Rapsheet: Continuous HTTP and DNS monitoring | Keagan Jarvis | 27m |
Rapsheet continuously runs 7,000 tests every 90 seconds across ~800 customer consoles, performing four DNS queries (including TXT ping) and five HTTP checks… | ||
| Talk Data to Me - the art of storytelling | Diana Pholo | 24m |
Shows a concrete pipeline: clean missing values, derive age from birthdate, and create features with Pandas, Matplotlib, Seaborn to inspect distributions… | ||
| Introduction to the PSSSA | David Sharpe | 05m |
A registered non‑profit in South Africa that supports and grows the Python community, handling core infrastructure and finances while giving grants and… | ||
| Friday Lightning Talks | 01h 20m | |
Lightning session hits concrete takeaways: Uvicorn + Starlette as a tiny ASGI web toolkit with WebSocket and GraphQL support and minimal dependencies… | ||
| There will be Data: Scraping the Web with Python | Andrew Collier | 36m |
Scrapes a static MPs directory to extract names, parties, emails and phone numbers using requests to fetch HTML and BeautifulSoup with CSS selectors (including… | ||
| Edge computing at the edge of the world | Schalk Heunis | 37m |
Builds a real-time music recommender that runs on Adafruit ItsyBitsy microcontrollers using CircuitPython, extracts song features by converting tracks to… | ||
| Migrating to Python 3 using Docker for your CI/CD pipeline (even if you can't use it in production) | Cobus Bernard | 32m |
Shows a concrete path to migrate a Python 2.7 app to Python 3.7 by building and testing inside containers, then using the container as a packaging artifact to… | ||
| My Journey into Buzzword Bingo | Sibulele Hlongwane | 26m |
A junior data scientist recounts jumping straight from a CS degree into practical work, discovering that most time goes to business understanding, data… | ||
| Launching a library(API client) in total darkness | Ronald Maravanyika | 34m |
Built a Python API client for Degreed using OAuth2 client‑credentials (client_id/secret + scope) and keeping sessions with requests.Session to handle high… | ||
| Thursday Lightning Talks | 30m | |
A rapid sequence of practical demos and claims: an inverted UI argument that 99% of apps force naming before intent and a proposed incremental UI pattern; a… | ||
| Making sense of Cape Town using NLP | Gordon Inggs | 39m |
City HR job descriptions become numeric: ~100,000 rows for ~15,000 positions are embedded with the spaCy large model into 300-dim vectors, reduced per person… | ||
| From data science to scalable NLU and vision cloud service. | Bernardt Duvenhage | 39m |
Built a private NLU and vision cloud service handling 5M+ queries/month, implemented in Python with NumPy, PyTorch, TensorFlow, OpenCV, TorchVision… | ||
| Closing Ceremony | 16m | |
Conference closes with concrete takeaways: Obsidian iPods giveaway and a Microsoft Antony backpack prize winner named, sponsors listed (Obsidian, Microsoft… | ||
| Opening Ceremony | 12m | |
Welcome and logistics rundown for PyConZA 2019 details sponsors and venue arrangements, plus what actually happens each day. Sponsors include Obsidian… | ||
| nothing matches | ||