Kenny Vaneetvelde’s Post

I’ve stumbled across Bee Agent Framework, and it’s been an impressive experience. If you’re exploring agentic AI workflows, this framework deserves your attention. What struck me the most is how intuitive and modular it is. Perfect for building intelligent agents that can handle complex tasks, and its almost production ready. Workflows felt natural, and has a clean architecture. Best of all it’s open-source, so there’s plenty of room to explore, adapt, and contribute to its community. Its OpenAI compatible or BYO(Model). Highly recommend giving it a try. npm install bee-agent-framework to get started Would love to hear your thoughts if you’ve used it! Let’s compare notes. 🐝 https://lnkd.in/eTQH7Zir #AI #AgentFramework #BeeAgentFramework #AIWorkflows #OpenSource #SmartAgents #AgenticAI

The Atomic Agents framework is designed to be modular, extensible, and easy to use. Its main goal is to eliminate redundant complexity, unnecessary abstractions, and hidden assumptions while still providing a flexible and powerful platform for building AI applications through atomicity. The framework provides a set of tools and agents that can be combined to create powerful applications. It is built on top of Instructor and leverages the power of Pydantic for data and schema validation and serialization. https://lnkd.in/dsEfXdGK

Boost your productivity by automating README file creation with GitHub Actions and AI (Gemini). Check out my latest guide to streamline your workflow! Link: https://lnkd.in/dpC3pjnJ #GitHub #AI #Automation #ReadMeGenerator"

💡 Exploring New Horizons in AI with the IBM Bee Agent Framework 🐝🤖 I'm thrilled to share my first steps contributing to the IBM Bee Agent framework! https://lnkd.in/degvcHaH This journey is an incredible opportunity to refine my understanding of agent-based systems, enhance interaction patterns, and explore new ways to build more effective AI tools. What I'm Working On: I’ve been experimenting with HumanTool and improving decision-making processes in agent interactions. My current draft already works for simple cases, and here's a sneak peek of a working example: Example Interaction: 👉 User: "Can you write the formula to calculate the area of a triangle?" 🤖 Agent: Provides a clear and accurate response. 👉 User: "I need help to calculate an area of a shape." 🤖 Agent: Realizes more details are needed, calls HumanTool to gather additional input, and then provides the formula for a circle based on user clarification. Why This Matters: 1️⃣ Learning: I'm diving deep into the Bee Agent framework to expand my understanding of agent orchestration and interaction. 2️⃣ Exploration: These experiments inspire me to innovate in creating seamless, intuitive interactions between users and AI agents. 3️⃣ Application: I'm beginning to migrate parts of Project Copilot, my AI Project Manager, to this framework to leverage its strengths for planning and backlog generation. Next Steps: I’m actively refining the prompts, enhancing modularity, and aligning my work with community best practices. Early feedback from this draft will help me iterate toward a cleaner, more robust implementation. 📢 I’d love to hear your thoughts or suggestions as I continue this exciting journey! If you’ve worked with the IBM Bee Agent framework or similar systems, I’d be delighted to connect and exchange ideas. Let’s build something great together! 🌟 #AI #OpenSource #BeeAgentFramework #GenerativeAI #ProjectCopilot

I’m excited to share a recent project where I utilized a Retrieval-Augmented Generation (RAG) system enhanced with a Generative Feedback Loop (GFL) to improve document retrieval and response generation in Chatbot. Key Components: 1- Document Processing: I leveraged LangChain open source framework to load and split a PDF document into chunks. These chunks are then embedded using Google's Generative AI model (Gemini). 2-Database Integration: The embedded chunks are stored in a ClickHouse database for quick retrieval (MYScaleDB). 3- RAG System: When a user asks a question, we find the most relevant document chunks using vector similarity search within the database. These relevant chunks are used to create a context-aware prompt for the Gemini Pro model. For more efficiency I've added an Agent performing Generative Feedback Loops (GFL) to The Retrieval Augmented Generation (RAG) pipeline to boost speed and reduce costs. GFL save LLM-generated results to the database, expanding the chatbot's knowledge base and improving query handling. This approach reduces costs by using cached results for similar queries. For Simplicity: When a user asks a question, we search past question/answer pairs for similar ones. If no similar question exists, we use an expensive model to generate a response with vector database context in my case I used Gemini. If similar questions exist, we use a cheaper, smaller model to adjust the response for this I used flan-t5-small. Here is the github repository: https://lnkd.in/d9icUAdC

#Topics How AI assistants are already changing the way code gets made [ad_1] The key idea behind Copilot and other programs like it, sometimes called code assistants, is to put the information that programmers need right next to the code they are writing. The tool tracks the code and comments (descriptions or notes written in natural language) in the file that a programmer is working on, as well as other files that it links to or that have been edited in the same project, and sends all this text to the large language model behind Copilot as a prompt. (GitHub co-developed Copilot's model, called Codex, with OpenAI. It is a large language model fine-tuned on code.) Copilot then predicts what the programmer is trying to do and suggests code to do it. This round trip between code and Codex happens multiple times a second, the prompt updating as the programmer types. At any moment, the programmer can accept what Copilot suggests by hitting the tab key, or ignore it and carry on typing.  The tab button seems to get hit a lot. A study of almost a million Copilot users published by GitHub and the consulting firm Keystone Strategy in June—a year after the tool’s general release—found that programmers accepted on average around 30% of its suggestions, according to GitHub’s ...

Another exciting development in the area of multi-agentic solution, details are in given link https://lnkd.in/dVt2QWbs and the GitHub repo - https://lnkd.in/d4wuMwzF (from same article) I am understanding difference of Swarm and Langframe given as below, share your views as well, is it fair to compare LangFrame and Swarm? I see differences given as below 1. Swarm: Swarm is more focused on lightweight, client-side multi-agent orchestration. It's built to easily create and manage multi-agent systems on the client side, ensuring seamless collaboration between agents. The agents in Swarm are designed to communicate dynamically and pass tasks to one another. Langframe: Langframe is designed as a more robust, server-side multi-agent framework. It integrates deeply with language models and allows for orchestrating not just agents, but tools, APIs, and workflows in a scalable environment. It’s more versatile in terms of backend scalability and is suited for larger, enterprise-level systems. 2. Swarm: Offers agent hand-offs, shared context variables, and tool execution, which means developers can create custom agents that work with specific data. The framework focuses more on agent-to-agent orchestration, making it highly modular for specific use cases. Langframe: While Langframe also supports custom agents, its strength lies in integrating multiple tools, workflows, and APIs. It’s designed to support complex workflows where agents need to interact with external systems (like databases, APIs, or external applications). #Swarm #MultiAgentSystem #OpenAI #Framework #Langraph

Agentic Frameworks: Key Features Comparison Here's what we know so far: The recent release of OpenAI's Swarm has let people wonder, What is the core difference between Swarm and other popular frameworks? Today, in the given diagram we breakdown few points for each framework. In my POV: 📌 OpenAI Swarm is ideal for developers looking to create complex multi-agent systems with a focus on real-time coordination and customization. 📌 LangGraph is suited for those who prefer visualizing workflows and managing intricate interactions among agents using a graph-based approach. Its replay feature makes it more advanced than others which maybe perfect for precise interactions. 📌 Autogen is best for users seeking a straightforward, open-source solution for simpler tasks, 📌 While CrewAI caters to organizations needing a structured, role-specific framework with user-friendly management tools. Same here, the replay feature makes it easier for your organizations to implement specific interactions. So each of them caters to their own specific use-cases! Additionally, Swarm is pretty new in contrast with others. So, updates might change a thing or two in the future! What are your thoughts on the new framework? Let me know in the comments below! Please make sure to, ♻️ Share 👍 React 💭 Comment to help more people learn

🐤 Canary Search & Ask AI across your docs(webpage), GitHub issues, and discussions. #AIPoweredSearch #CanaryTool #DocSearch #GitHubIntegration #ArtificialIntelligence #SoftwareDevelopmentTools #CodeDiscovery #KnowledgeManagement #DeveloperProductivity #AIassistedDevelopment https://lnkd.in/gCj3Kw6x

Great to see the UK government and the Incubator for Artificial Intelligence open source the code for the new government Gen.AI Redbox Copilot. Having actively trialed it within three different Private Offices with the Cabinet Office, the team is now broadening its scope to address the entire civil service. This a great resource to show an example of a working Gen.AI app which includes features such as: ☑ Retrieval Augment Generation (RAG) based chat for data-driven conversations with granular citations. ☑ Guided summarisation, where the human in the loop can review and guide summaries of multiple documents. ☑ Advanced prompting techniques to ensure safety and balance in AI responses and outputs. ☑ Proactive data privacy and retention features to maintain compliance and manage data. ☑ Securely designed for on-premises deployment and across clouds. The repo is at https://lnkd.in/gw4SVwum #llm #generativeai #largelanguagemodels