Onya: an open source #KnowledgeGraph system with a literate bent for the #LLM world, based on my Versa work. In all my experience with #LLMOps, it benefits greatly from a more literate approach to supporting tech, hence Word Loom, Amara & Onya.
github.com/OoriData/Onya/
#Python #LLMstack #GenAI

Agentic Frameworks on rise and will continue to see more in 2025 (mostly Python)👇
1️⃣ Microsoft AutoGen
2️⃣ LangGraph / LangChain
3️⃣ AutoGPT
4️⃣ CrewAI
5️⃣ Swarm by OpenAI
6️⃣ MetaGPT
7️⃣ Flowise
8️⃣ OpenAGI
9️⃣ HF Transformers Agents
🔟 Camel

Others: Langflow, LLMStack

Are you aware of any more?

#GenAI #Python

New Onya release, 0.2.0. Includes Mermaid and graphviz output. Onya is a #KnowledgeGraph format and toolkit with a literate bent for the #LLM world, based on my long-standing Versa work.
#Python #LLMstack #GenAI
github.com/OoriData/Onya

Next up in this interlocking series of updates for my #FLOSS data processing & #GenAI workflow projects: Amara. Around since the mid 2000s in various iterations, it's been re-consolidated & updated. URL/URI/IRI, #XML, #XPath etc.
github.com/OoriData/Amara
pypi.org/project/Amara/
#Python #LLMstack

#LLMCounsel #AIArchitecture #MultiModelAI #LLMStack #Deliberation

https://github.com/trypromptly/LLMStack
more like this: https://agentwire.kynth.studio/?utm_source=agentwire-bluesky&utm_medium=social

LLMStack is the no code option: chain multiple LLM agents into workflows and apps over your own data, without wiring the orchestration by hand. From trypromptly. #dev

https://github.com/trypromptly/LLMStack

https://github.com/trypromptly/LLMStack/

Cognee isn't just embedding chunks.
It creates connections + indexes + meaning.
Perfect for RAG setups with custom DBs.
Feels like bridging data and cognition.

#Cognee #RAG #LLMStack
#LLmZoomcamp

Claude helps you reason.
GPT-4o helps you write.
Gemini simplifies.

But only Kynismos lets them work together — privately — for you.

🧠 One chat. Infinite power.

www.kynismos.ai

#MultiAI #BoundlessPotential #LLMstack #aitips

LangChain feels like overkill until it’s not.
The second you want your AI to talk to files, remember things, or make decisions based on user input, you’re going to wish you started learning it last week.
#langchain #llmstack #openai

2/ The model is just the engine. The real value comes from what wraps around it: context management, workflows, memory, tool integration, and intuitive UX. That’s where the difference is made.
#AIInfrastructure #ProductDesign #LLMStack