Rapid-MLX

Run AI on your Mac. Faster than anything else.

https://github.com/raullenchai/Rapid-MLX

How we rebuilt search ranking at Faire with deep learning

From XGBoost to deep learning: a two-year rebuild of Faire’s ranking stack.

https://craft.faire.com/how-we-rebuilt-search-ranking-at-faire-with-deep-learning-14f080679c83

Full-Text Search with DuckDB

The post shows how DuckDB’s full-text search extension can index a large email corpus and run BM25-ranked keyword search directly in SQL, without needing a separate search engine. It also walks through practical preprocessing and filtering steps, then demonstrates conjunctive queries that return only documents matching all search terms.

https://peterdohertys.website/blog-posts/full-text-search-w-duckdb.html

lightning PyPI Compromise: A Bun-Based Credential Stealer in Python

The post describes a PyPI supply-chain compromise in lightning 2.6.2/2.6.3, where importing the package silently downloads Bun and runs an obfuscated JavaScript credential stealer. It also says the payload steals GitHub, cloud, and other secrets, then uses any captured credentials to spread further and commit exfiltrated data back into victim repos.

https://snyk.io/blog/lightning-pypi-compromise-bun-based-credential-stealer/

token-optimizer

Find the ghost tokens. Fix them. Survive compaction. Avoid context quality decay.

https://github.com/alexgreensh/token-optimizer

foundry

Ship full-stack agentic systems the way they're meant to be built - production-ready, secure by default, with the developer experience modern Python deserves.

https://github.com/promptise-com/foundry

z-lab / dflash

DFlash: Block Diffusion for Flash Speculative Decoding

https://github.com/z-lab/dflash

Using Bag-of-Words With PyCharm

Let's unpack how the bag-of-words model works, explore the techniques it uses to convert text into numerical representations, and look at where it fits relative to more modern NLP approaches.

https://blog.jetbrains.com/pycharm/2026/04/using-bag-of-words-with-pycharm/

bluerock

Runtime visibility for Python MCP servers. Captures tool calls, session lifecycle, module imports (SHA-256), and subprocess execution as structured NDJSON. No code changes. Apache 2.0

https://github.com/bluerock-io/bluerock

Fast Mesh Booleans in Python

Learn how to perform fast mesh boolean operations in Python. Union, intersection, and difference at interactive speed on million-polygon meshes. One pip install, NumPy arrays in and out.

https://polydera.com/tutorials/fast-mesh-booleans-in-python

Boosting multimodal inference performance by >10% with a single Python dictionary

Multimodal models are promising, but inference engines haven't been optimized for them yet. We profiled SGLang’s scheduler on a multimodal workload and identified an opportunity to replace expensive book-keeping around shared GPU memory with a simple cache lookup. Throughput and latency both improved over 10% on our target workload.

https://modal.com/blog/boosting-multimodal-inference-performance-by-greater-than-10-with-a-single-python-dictionary

auto-round

A SOTA quantization algorithm for high-accuracy low-bit LLM inference, seamlessly optimized for CPU/XPU/CUDA, with multi-datatype support and full compatibility with vLLM, SGLang, and Transformers.

https://github.com/intel/auto-round