Q: Connect to Academic Sources
Hi.
Congratulations on your campaign. I am interested in purchasing the AgenticFlow T4, but I need to confirm a few things first:
1. Can AF connect to academic sources such as Scopus API, Google Scholar, or Semantic Scholar? Or does it have a native connection with applications like Afforai (It was on Appsumo last year)?
2. Is there an add-on for knowledge storage if our storage becomes full in the future?
SeanP_AgenticFlowAI
May 11, 2025A: Hey Galih52553!
Thanks so much for the congrats and your interest in Tier 4! Great questions.
1. Connecting to Academic Sources (Scopus, Google Scholar, Semantic Scholar, Afforai):
- Direct MCPs: Currently, we don't have pre-built, dedicated MCP connectors specifically for Scopus API, Google Scholar (which has very limited/no official API for direct content scraping by third parties), Semantic Scholar, or Afforai.
- Your Roadmap Post - Thank You!: I saw your excellent feature request on our roadmap for this: https://agenticflow.featurebase.app/p/access-to-academic-sources-database. That's exactly the right way to get these prioritized! The more votes, the higher it goes on our list for potential native MCP development.
- Current Approach (API Call Node): If Scopus, Semantic Scholar, or Afforai offer REST APIs, you can absolutely connect to them right now using AgenticFlow's API Call node within a workflow. You'd configure it with their API endpoint, any necessary authentication (API keys, etc.), and the query parameters. Your workflow would then process the JSON/data returned.
- Web Scraping (Caution): For sources like Google Scholar that lack robust APIs, you could attempt to use web scraping (e.g., via our built-in scraper or an Apify MCP), but this can be unreliable due to anti-scraping measures.
2. Knowledge Storage Add-on & Clarification:
- Not Just File Size: Just to clarify when we talk about knowledge storage (e.g., the 2GB on Tier 4, which maps to Team), it's not just raw file size. It refers to the storage needed for the vectorized embeddings of your knowledge. This is the "actual knowledge" the agent uses for semantic search (RAG). A 1GB vector store can hold a very significant amount of text information (tens of thousands of text chunks). So, the 2GB on T4 is indeed a lot of processed knowledge.
- Yes, Knowledge Storage Add-ons Are Available: If you do find you need more vectorized knowledge storage beyond your plan's limit in the future, you can purchase additional knowledge capacity. As per our pricing page (https://agenticflow.ai/pricing), it's typically around $100 per additional GB of vectorized knowledge storage.
Thanks for adding that to the roadmap – it's super helpful for us to see what integrations the community needs most!
— Sean