November 11, 2025 • A tutorial on building a stateful RAG agent with LangGraph for orchestration and Beanis for persistent memory and vector search, all backed by Redis, in about 200 lines of Python.
python langgraph beanis redis rag vector-search
November 05, 2025 • Concept-pedia is a wide-coverage, semantically-annotated multimodal dataset of 165K+ visual concepts built on BabelNet, with the Concept-10k benchmark and SigLIP models on Hugging Face for evaluating vision-language models beyond ImageNet.
conceptpedia multimodal vision-language emnlp2025 research dataset benchmark huggingface siglip babelnet
October 30, 2025 • How to use Redis (via the Beanis ODM) as a geo-spatial cache in front of PostgreSQL/PostGIS, with OpenStreetMap data, to serve nearby-restaurant queries in about 12ms.
python redis geospatial cache beanis tutorial architecture
October 27, 2025 • Word Sense Linking (WSL) is a model that finds the ambiguous words in raw text and links each to its correct WordNet sense, no pre-marked spans or candidate lists required. Published at ACL 2024 (Findings) by Babelscape and Sapienza. Includes the model, results, and Python usage.
nlp word-sense-disambiguation acl2024 research python transformers
October 24, 2025 • Build a RAG pipeline in about 50 lines of Python using Redis vector search and the Beanis ODM, without adding a separate vector database like Pinecone.
python redis RAG vector-search semantic-search beanis embeddings
October 23, 2025 • Beanis is an async, Pydantic v2 object-document mapper for vanilla Redis. It cuts serialization boilerplate by about 70% and stays within roughly 8% of raw redis-py performance.
python redis database odm pydantic open-source