For AI products

Ground your AI assistant in real scholarships.

Language models are confident about scholarships that don't exist and deadlines that have passed. Give yours a catalog to answer from, with the funder's own words and a link to apply.

Talk to us See Catalog API

How do I stop my AI assistant making up scholarships?

Don't let it answer from memory. Sync a verified catalog like TuitionCovered's into your own store, give the model a search tool over it, and tell it to recommend only what the tool returns, with the official link. Unknown fields are null in our data, so there's nothing for the model to mistake for a fact.

What you can build

“What can I apply for?”

Retrieve from your synced copy with the student's state, level and GPA, and answer with real awards.

Answers with sources

Quote eligibility in the funder's own words and link to the official page, so students can check.

Deadline-aware advice

Only suggest what's still open, and say “month only” when that's all the funder gave.

A hand-off to apply

End the conversation with apply_url, and see which suggestions turn into clicks.

How it works

Sync the catalog into your store, then expose a search tool like this one to the model.

1
Sync
Store the catalog in your database or vector store, refreshed with updated_since.
2
Expose a tool
Give the model a search tool over your copy, like the definition below.
3
Constrain
Instruct it to recommend only tool results, with the official link.
4
Measure
Watch apply clicks per scholarship in the portal.

A tool for your assistant

{
  "name": "find_scholarships",
  "description": "Search open scholarships the student may qualify for. Only recommend results from this tool, always with official_url; never invent scholarships or deadlines.",
  "input_schema": {
    "type": "object",
    "properties": {
      "state": { "type": "string", "description": "USPS code, e.g. OH" },
      "education_level": { "type": "string", "description": "e.g. HS_12, UG_1" },
      "gpa": { "type": "number" },
      "first_generation": { "type": "boolean" },
      "no_essay": { "type": "boolean" }
    }
  }
}

Fields you'll use

Every field is always present in the record, null when the funder doesn't say.

Field	What it's for
eligibility.criteria	The funder's own wording, for the model to quote
eligibility.* codes	Structured filters for the tool's arguments
official_url	The source to cite
dates.deadline, dates.deadline_precision	Only suggest open awards, at the precision given
quality.last_verified_at	When the record was last confirmed
apply_url	The hand-off, credited to your company

Every field in the API reference

Why TuitionCovered

Structured eligibility that maps cleanly to tool calls and filters
Eligibility criteria as the funder wrote them, for quoting
Unknown means null: nothing for the model to mistake for a fact
Incremental sync keeps your index current with updated_since

What to use

AI assistants: common questions

Does the API do the search for me?

The API serves the whole active catalog, with incremental sync. Keep a copy in your own database and filter it there by state, level, GPA or anything else in the record.

Can I put the data in a vector database?

Yes, your contract permitting. Embed the summary and eligibility criteria for retrieval, and keep the structured fields for filtering.

How do I keep answers current?

Re-sync daily with updated_since and drop any record that stops being served. Each record's last_verified_at says when it was last confirmed.

What does it cost?

It depends on your traffic and how you'll use the data. Tell us what you're building and roughly how many students you expect, and we'll quote. You start on a pilot with real keys before you sign.

All use cases

Tell us what you're building

We'll show you the data for your audience and set you up with the right option.

Talk to us