OpenContext - Give Your AI Assistant a Persistent Memory for Multi-Agent Workflows
A personal context store for AI agents: CLI + MCP + Desktop app to capture, search, and reuse project knowledge across agents and repos
Fodev JEO
LATEST INVESTIGATIONS
Source audits & field notes
Rnj-1 and Motif-2-12.7B: Why These Open-Source Models Are Reshaping the LLM Landscape
Deep dive into Essential AI’s Rnj-1 (8B) and Motif Technologies’ Motif-2-12.7B - two open-source models that challenge the status quo with exceptional performance
All Agentic Architectures - A Comprehensive Guide to 17+ State-of-the-Art AI Agent Patterns
A deep dive into the all-agentic-architectures repository covering 17+ production-ready agentic AI patterns from Reflection to Reflexive Metacognitive agents, built with LangChain and LangGraph
The Production Generative AI Stack - Architecture and Components
Exploring the modern AI technology stack that powers production-grade generative AI applications, from accelerated compute to user experience layers
Evaluating DeepAgents CLI on Terminal Bench 2.0
How DeepAgents CLI performs on Terminal Bench 2.0, achieving ~42.5% accuracy with Claude Sonnet 4.5, and the infrastructure needed for isolated agent evaluation
HuggingFace Text Embeddings Inference (TEI) - A Deep Dive into High-Performance Embedding Architecture
Exploring how TEI combines Rust’s performance with Python’s flexibility to deliver blazing-fast text embeddings inference, from architecture to production deployment
Combining Rust and Python for High-Performance AI Systems - Bridging Speed and Productivity
Exploring how Rust and Python complement each other in AI systems, overcoming Python’s performance bottlenecks while maintaining its rich ecosystem
NVIDIA CUDA Tile - Focus on Your Algorithm, Let Hardware Handle the Rest
Exploring CUDA Tile, NVIDIA’s revolutionary tile-based programming model that abstracts tensor core complexity, enabling higher-level GPU programming without sacrificing performance
Writing Blogs with Multi-Agent Systems - A Production Framework for AI-Powered Content Creation
Exploring how multi-agent systems can revolutionize blog writing workflows, featuring KTransformers for efficient LLM inference and heterogeneous computing
Memori: SQL-Native Memory Engine for LLMs and Multi-Agent Systems
Exploring Memori’s open-source approach to persistent, queryable memory for AI agents using standard SQL databases - enabling 80-90% cost savings while maintaining context across sessions
Search article titles, categories, and tags. Full text is searched when needed.
Type to search articles.