Avi Avni

Token costs spiking. Responses too slow. Users don't trust the answers. All three are retrieval problems. GraphRAG SDK 1.0 is out. Ranked #1 on GraphRAG-Bench against 8 systems. Fewer LLM calls, grounded answers, predictable cost. Open-source: github.com/FalkorDB/Graph…

29 APRIL: Learn how to combine @cocoindex_io 's incremental data transformation capabilities with FalkorDB's graph and vector search to build scalable, low-latency GraphRAG systems that keep AI models grounded in fresh, structured knowledge. luma.com/4rlvnwx0 10:00 am PST

Your AI coding assistant now knows FalkorDB. Say good-bye to: Wrong Cypher syntax, missing index types or unsupported clauses. These errors cost you debugging hours that never needed to exist - as well as costly tokens! Get the skills here: github.com/FalkorDB/skills

NEW: FalkorDB now runs as a @Snowflake Native App 🏂❄️ You bind your tables, load data into a graph, and write Cypher queries: all inside your Snowflake account. No data movement or external infrastructure -> app.snowflake.com/marketplace/li…

SAVE THE DATE: Join us for a technical session on advanced graph algorithms, with a cyber-security use case, user-defined-functions, and the recently-launched FLEX library by FalkorDB's CTO and Solution Architecture team. 17 + 18 March, 2026 (two sessions), links below

🚀 FalkorDB MCP Server v1.1.0 is live! Connect your AI agents directly to a graph database using the Model Context Protocol. Works with Claude, Cursor, and any MCP-compatible client — out of the box. Docker-ready. Open source. MIT licensed. 👉 github.com/FalkorDB/Falko…

Thanks Gartner for recognizing FalkorDB as a leading GraphRAG vendor! If you’re building AI applications, combining Knowledge Graphs with LLMs is today’s smartest strategy. GraphRAG enhances retrieval performance by 35–46%, especially for complex, multihop, and temporal queries

@g_korland @DvirDu Why no assembly or binary? The llm write perfect code without feedback? Why the llm not just generate pixels no need computation at all?

@DvirDu I said "perfect code", but soon anyway most of the code it will learn from will be generated by other Agents

Assuming Code Agents are going to write a perfect code soon enough, why would we need Rust? Or any other server side language with some overhead like Java or Go (let alone nodejs)? Why shouldn't perfect agents just write super efficient C code?

Checking my open source footprint. 🌍 ⚡️ GitPulse Score: 99/100 🏆 Rank: Exceptional Output Check my profile on gitpulse.dev: gitpulse.dev/profile/aviavn…

Security data is inherently Graph-shaped. Security data is graph-shaped. Incidents span identity, workload, cloud & network, all highly interconnected. We're coming to present this & more at RSAC 2026. If you'd like to learn more, let's connect #RSAC

🎉Drumroll, please!🎉 The top startups of the week! 💯💡🌟 📊@falkordb 🚀@wearetheswarm 🌐@LinkcardD 🛍️@SekelTech 👶@tinylaneindia Would you like your startup to be the next one featured? Apply for FREE 👉 bit.ly/3mKVo1u #InnovativeStartups #Entrepreneurship

#MCP for agent memory: @zep_ai Graphiti + FalkorDB for persistent, multi-tenant knowledge graphs. youtu.be/tjEznowIhBo

FalkorDB is now trending this month, right alongside legendary projects like Git!

Meet me at AI Dev TLV! Can't wait to talk AI, Agentic Memory and Graph Databases. See you there! #AgenticAI

BREAKING: Your AI agents have the memory of a goldfish. They struggle with both short-term and long-term memory, so users have to keep reminding them who they are, it feels like talking to an agent with amnesia. FalkorDB | @falkordb is the open-source, Redis-module graph database, and made to be the perfect long-term memory for agents. It keeps real relationships between facts, so GraphRAG answers stay sharp where VectorRAG gives you similarity, not relevance. Why devs love it: → 10,000+ graphs in one database → Multi-tenancy on every tier → Cloud or your own servers in minutes → Free tier with enterprise features Ultra-fast, grows with your data, and built for agent memory optimization. Try it now 👉 tryit.cc/vrkWS9 Star the repo → github.com/FalkorDB/falko…

Most vector databases are forgetting what makes your data valuable. They store chunks — not context. @falkordb brings relationships back to Generative AI. The world’s fastest open-source graph database for agentic memory.

Love seeing what people are already building on @Railway with @FalkorDB, like this highly available @n8n_io setup! Try it on Railaway railway.com/new/template/f…

Catch @g_korland's @Redis + Graph = Agentic Memory’s Secret Sauce talk on 13 Nov at Redis Released! Joining a great cohort of Redis, @Payoneer and Global E execs to discuss fewer hallucinations, latest techniques, and actually reaching production. eu1.hubs.ly/H0p86fR0

🔴 You're invited: Implementing Agentic Memory workshop (featuring @zep_ai #Graphiti) Tomorrow (9.9.2025). Sign up: lu.ma/c7u104z5 #AIagents #AgentMemory