If you’re reading this, chances are you’re in the same boat I was just a few days ago: staring at a growing list of agentic AI frameworks, wondering which one to learn first.
I had three exciting AI automation projects brewing in my mind:
- A personalised email outreach system for a company hackathon
- A SQL agent which generates context-aware queries and runs them to provide insights to automate my day job
- An intelligent document processing pipeline for a client
But here’s the thing—I was stuck before I even started coding.
The Framework Paralysis Problem
The agentic AI space is exploding with options. Every week, there’s a new framework promising to be the “ultimate solution” for AI automation. After weeks of research, I narrowed it down to four popular contenders:
n8n – The Visual Workflow Builder
“Perfect for non-coders,” they said.
- Visual, drag-and-drop interface
- Extensive integrations with popular services
- Great for business process automation
LangGraph – The Complex Reasoning Champion
“Build stateful, multi-actor applications,” they promised.
- Graph-based architecture for complex workflows
- Deep integration with the LangChain ecosystem
- Ideal for research and sophisticated reasoning tasks
CrewAI – The Role-Playing Specialist
“Orchestrate autonomous AI agents,” they claimed.
- Role-based agent system
- Collaborative workflows between different AI personas
- Perfect for creative and team-simulation projects
AutoGen – The Conversational Powerhouse
“Multi-agent conversation systems,” Microsoft declared.
- Advanced conversational AI capabilities
- Human-in-the-loop workflows
- Excellent for interactive and code generation tasks
The Problem: Which One First?
Each framework had compelling use cases. Each had passionate communities singing their praises. But I had a limited amount of time and mental bandwidth to learn something new.
The questions haunting me:
- Which framework best suits my project’s specific requirements?
- Should I consider my coding experience level?
- What are the hidden complexities I might face?
- Which one has the best learning curve for my background?
So instead of picking a framework randomly, I decided to build an AI-powered recommendation engine that could:
✅ Analyse project requirements intelligently
✅ Consider the user’s technical background
✅ Provide personalised recommendations with confidence scores
✅ Offer implementation tips and potential challenges upfront
✅ Give alternative options for different scenarios
Building the Solution
I spent an hour prompting Claude to create me an AI Framework Finder using Streamlit and Google’s Gemini AI. The tool works by:
- Smart Validation: First, it checks if your request is suitable for agentic AI systems
- Requirement Analysis: It analyses your project description using AI to understand complexity, integration needs, and automation requirements
- Experience Matching: It factors in your coding experience, crucial for framework selection
- Intelligent Recommendation: Using all this data, it provides a personalised recommendation with detailed reasoning
- Comprehensive Guidance: Beyond just picking a framework, it offers implementation tips and warns about potential challenges
Try the AI Framework Finder and let me know what it recommends for your project
Leave a Reply