A pandas merge can silently multiply rows or match missing keys. Learn how to enforce join cardinality, audit unmatched records, and verify that enrichment preserves the meaning of your dataset.
A hands-on 2026 guide to Ollama 0.34 in Python: install the runtime, wire up the ollama client, force JSON-schema outputs with Pydantic, call tools, generate embeddings, and stand up a fully local RAG pipeline against pgvector, with llama.cpp, LM Studio, and vLLM compared.
Prefect 3.8 in Python 2026: hands-on guide to flows, deployments, work pools, automations, and transactions for reliable data pipelines.
A hands-on tour of Apache Airflow 3.3 with the airflow.sdk Task SDK: TaskFlow decorators, task groups, dynamic task mapping with .expand(), assets, native DAG versioning, and a 2.x migration playbook drawn from moving 900 DAGs.
A FastAPI developer's tour of Modal in 2026: decorators, per-second GPU pricing, Volumes, @app.cls with @modal.enter, FastAPI endpoints, spawn_map for batch fan-out, and the cold-start tuning knobs that decide whether Modal saves money or bleeds it.
Zarr-Python 3 brings full v3 spec support, an async core, and chunk sharding for cloud object stores. A data-engineering walkthrough with chunking rules, migration steps, and pipeline tests you can actually run.
Benchmark Cohere Rerank 3.5, BGE v2-m3, Jina Reranker v2, and ColBERT v2 for RAG in Python. Runnable code, NDCG@10 results, latency, and $/1M queries so you can pick the right reranker.
DataFusion is an Apache Arrow-native, Rust query engine you install via pip as datafusion-python. Learn install, SQL and DataFrame APIs, UDFs, Substrait, Ballista, and how it stacks up against DuckDB and Polars in 2026.
uv is Astral's Rust-based Python package manager that replaces pip, pip-tools, pyenv, pipx, and Poetry with one tool that resolves and installs dependencies 10-100x faster. This 2026 guide covers uv.lock, PEP 723 scripts, workspaces, PyTorch/CUDA installs, and Jupyter integration.
A hands-on 2026 walkthrough of PySpark 4.0: Spark Connect from your IDE, custom data sources in pure Python, ANSI mode by default, VARIANT, polymorphic UDTFs, and how to migrate an existing Spark 3.5 pipeline without breaking overnight jobs.
DuckLake 1.0 keeps table metadata in a real SQL database and data as Parquet on S3. A practical Python guide covering pyducklake, time travel, deployment, and where it beats Iceberg.
A pragmatic 2026 guide to data contracts in Python: define them with datacontract-cli, enforce with dbt model contracts and Pandera, and wire them into CI so breaking changes fail the PR, not the 3am pager.
Choose your preferred language to explore our content
