Today's RAG & Vector Databases: Fastest-Growing Projects — August 30, 2026
Today's the RAG & Vector Databases space, we see a mix of established and emerging projects making waves on GitHub. Projects like deeplethe/utopia continue to gain traction with their innovative approaches to knowledge graph integration for Retrieval-Augmented Generation systems. Meanwhile, new entrants such as parqdb-io/parqdb are drawing attention with their scalable vector database solutions built entirely on Parquet and Arrow.
deeplethe/utopia
Self-hosted knowledge platform Utopia leverages Rust and PostgreSQL to provide a RAG system over documents integrated into a knowledge graph, ensuring temporal context for facts. With its high growth score of 28.59 and nearly 700 stars, Utopia is rapidly gaining recognition for its unique approach to managing evolving factual information within a structured database environment.
parqdb-io/parqdb
ParqDB offers an embedded vector database capable of handling billions of records efficiently by leveraging Parquet and Arrow technologies. Its strong growth score of 11.43, coupled with steady development activity (91 commits in the last month), highlights its growing importance in the realm of high-performance, scalable storage solutions for large-scale vector data.
makralabs/makra
Makra serves as a memory layer between web content and AI agents, delivering real-time structured data via vector search. With a growth score of 10.05 and consistent development efforts (30 commits in the last month), Makra is gaining traction for its potential to enhance real-time information retrieval and processing capabilities for AI applications.
jsdhwfmax/EvalForge
EvalForge provides an open-source platform aimed at evaluating the quality and security of RAG applications and AI assistants. With a growth score of 8.42 and modest development activity (11 commits in the last month), EvalForge is attracting interest from developers seeking standardized evaluation tools for their RAG systems.
TOPDEV99999/ai-ShopMind
Ai-ShopMind converts product descriptions into vector embeddings stored in Endee, a high-performance vector database. It then uses these embeddings to find semantically similar products based on customer queries and generates relevant answers grounded in the actual product catalog. Despite having fewer stars (29) and minimal recent activity (1 commit), Ai-ShopMind demonstrates potential in the e-commerce space by combining advanced AI techniques with practical use cases for retail businesses.
These projects illustrate a dynamic ecosystem where developers are pushing the boundaries of vector databases and RAG systems, driving innovation through both novel approaches and robust technical implementations.
deeplethe/utopia
Self-hosted knowledge platform Utopia leverages Rust and PostgreSQL to provide a RAG system over documents integrated into a knowledge graph, ensuring temporal context for facts. With its high growth score of 28.59 and nearly 700 stars, Utopia is rapidly gaining recognition for its unique approach to managing evolving factual information within a structured database environment.
parqdb-io/parqdb
ParqDB offers an embedded vector database capable of handling billions of records efficiently by leveraging Parquet and Arrow technologies. Its strong growth score of 11.43, coupled with steady development activity (91 commits in the last month), highlights its growing importance in the realm of high-performance, scalable storage solutions for large-scale vector data.
makralabs/makra
Makra serves as a memory layer between web content and AI agents, delivering real-time structured data via vector search. With a growth score of 10.05 and consistent development efforts (30 commits in the last month), Makra is gaining traction for its potential to enhance real-time information retrieval and processing capabilities for AI applications.
jsdhwfmax/EvalForge
EvalForge provides an open-source platform aimed at evaluating the quality and security of RAG applications and AI assistants. With a growth score of 8.42 and modest development activity (11 commits in the last month), EvalForge is attracting interest from developers seeking standardized evaluation tools for their RAG systems.
TOPDEV99999/ai-ShopMind
Ai-ShopMind converts product descriptions into vector embeddings stored in Endee, a high-performance vector database. It then uses these embeddings to find semantically similar products based on customer queries and generates relevant answers grounded in the actual product catalog. Despite having fewer stars (29) and minimal recent activity (1 commit), Ai-ShopMind demonstrates potential in the e-commerce space by combining advanced AI techniques with practical use cases for retail businesses.
These projects illustrate a dynamic ecosystem where developers are pushing the boundaries of vector databases and RAG systems, driving innovation through both novel approaches and robust technical implementations.