Conf42 Machine Learning 2023
2023-05-18
-- online --
[Conf42 Machine Learning 2021, Conf42 Machine Learning 2022, Conf42 Machine Learning 2023, Conf42 Machine Learning 2025, Conf42 Machine Learning 2026]
who it's for
MLOps engineers, machine learning engineers, data engineers, data scientists, ML platform engineers, site reliability engineers
topics
edge MLOps and deployment; Apache Airflow pipelines; WebAssembly in-database UDFs; Polars dataframe optimization; cascade computer vision models; LLM reasoning techniques; BERTopic topic modeling; generative multimodal interfaces; vector databases and ops; production vector embeddings; experiment tracking and model registry; ML event streaming with Java
| title | author | length |
|---|---|---|
| Premiere | Noa Goldman, Alejandro Cantos, Tamara Janina Fingerlin, Akmal Chaudhri, Matt Harrison, Argo Saakyan, Logesh Kumar Umapathi, Abhiram Ravikumar, Jaspal Singh Jhass, Gaurab Patra, Yogesh Seenichamy, Daniel Svonava, Zain Hasan, David Kjerrumgaard | 08m |
A rapid lineup of short sessions promises concrete takeaways: why Edge is the new Cloud with cases that must not send data to the cloud, an Apache Airflow… | ||
| A Hitchhiker's Guide to the MLOps Experience | Noa Goldman | 24m |
Argues that applying software development practices to ML — specifically data versioning so datasets are reproducible like code, side‑by‑side visualizations… | ||
| Introduction to MLOps at the Edge | Alejandro Cantos | 47m |
Claim: the cloud is "broken" for latency, cost, vendor lock-in, privacy and eight other reasons, so compute moves closer to sensors. Demonstrations use Barbara… | ||
| Orchestrating data and ML workflows with Apache Airflow | Tamara Janina Fingerlin | 49m |
Pipeline built in Apache Airflow with eight DAGs and dataset-driven scheduling, using S3 for images, DocDB for metadata, Hugging Face ResNet50 fine-tuned via a… | ||
| Using WebAssembly for in-database Machine Learning | Akmal Chaudhri | 18m |
Builds a Rust WebAssembly module from a Rust port of Vader sentiment, compiles and deploys it into SingleStore DB with the push wasm tool, and exposes it as a… | ||
| Polars: A highly optimized dataframe library | Matt Harrison | 21m |
Polars 0.17 loads a 41k×83 CSV with scan_csv lazy + collect and through targeted casts and categoricals reduces memory from ~33 MB to ~2.5 MB (examples… | ||
| Cascade models in computer vision to boost accuracy and performance | Argo Saakyan | 12m |
Two-stage pipeline uses a fast high-recall detector on full frames downscaled to 640×640 to propose candidates, then a second-stage classifier trained on… | ||
| Unlocking reasoning in Large language models | Logesh Kumar Umapathi | 39m |
Survey of methods that reliably improve multi-step answers: chain-of-thought prompting and self-consistency boost accuracy on math and commonsense benchmarks… | ||
| Unveiling Clustering in BERTopic Topic Modeling | Abhiram Ravikumar, Jaspal Singh | 27m |
BERTopic is shown end-to-end on Amazon Alexa reviews from Kaggle using sentence-transformer embeddings, UMAP for reduction, clustering with HDBSCAN, and topic… | ||
| Generative AI will enable crowdsourced interactivity | Gaurab Patra | 30m |
Claims that combining GANs and Transformers produces true multimodal, interactive content and cites a platform context with almost 20 million swipes to show… | ||
| Utilizing The Power of Machine Learning in Healthcare | Yogesh Seenichamy | 16m |
Uses the Kaggle Heart Failure Prediction Dataset and a Colab notebook with TensorFlow/Keras to train a 3-layer neural net (input → 16 → 16 → 1) using one‑hot… | ||
| Vector Ops: run vector embedding-powered apps in prod | Daniel Svonava | 50m |
Argues vectors capture nuance lost to words and shows a working pattern: concatenate content and collaborative features into single vectors, store user and… | ||
| A gentle introduction to Vector Databases | Zain Hasan | 45m |
Explains that vector databases store embeddings plus original objects, use approximate nearest neighbor indexes to search billions of vectors, and support CRUD… | ||
| Build ML Enhanced Event Streaming Apps with Java Microservices | David Kjerrumgaard | 27m |
Demonstrates a runnable Java microservice pipeline that ingests simulated tweets, runs a bundled NLP model inside a Pulsar Function, and writes labeled results… | ||
| nothing matches | ||