I'm in a dilemma. Hey there, Hope everyone is doing good! 👋🏾
I’m working on building a scholarship research agent that can:
- Search and rank scholarships based on specific eligibility criteria like field of study, nationality, etc using live, real-time web data
- Store and reference results through Airtable
I’m currently weighing two approaches:
- using Claude with MCP (since it offers more integrated tool use) or
- building an n8n workflow using ChatGPT.
I already pay for ChatGPT Pro, so I'm not really looking to start paying another subscription for Claude unless it's better.
Questions
- Is it still worth exploring the n8n workflow with the ChatGPT route if the AI agent cannot be accessed directly in ChatGPT?
- If so, how would I package it in a user-friendly way for a client example?
- Since ChatGPT now has a browsing feature that can scrape real-time data from websites, is there any point in building an external agent at all?
- Claude’s free plan has prompt limits so would it even make sense to migrate the solution to Claude long term?
Any advise would be great.