Prototype demo

grantAgent

Weekly federal funding matches for tenure-track faculty — matched to each person's own publication record, with an explanation of why each opportunity fits.

Try the faculty view Open the admin console

You are viewing the site as Guimu Guo (Computer Science, Rowan University), using his real publication profile. The 161 opportunities are real grants.gov listings; colleagues shown elsewhere are fictional.

For faculty

Weekly matches→

Every Monday, each faculty member gets the federal opportunities that fit their research, each with a written rationale and a fit verdict.

Browse & save→

Search all open opportunities, sorted by how well they match your publication profile, deadline, or award size. Save the ones to come back to.

Publication profile→

Built automatically from OpenAlex. Faculty can remove papers that were misattributed to them, and the profile rebuilds.

For research office / admins

Faculty roster→

Who is subscribed, profile coverage, paper counts.

Match review→

Every match the pipeline produced, week by week, with retrieval and LLM scores.

Review queue→

Ambiguous author identities and suspected misattributed papers, approved or rejected by a human.

Agent runs→

Full transcripts of each AI agent's tool calls and verdicts, for auditing.

Cost & quality→

Weekly LLM spend per agent, match quality, and data-source health.

How it works

Step 1
Ingest
New federal opportunities are pulled from grants.gov daily, with full synopsis and eligibility.
Step 2
Profile
Each faculty member's research profile is built from their publications (OpenAlex / ORCID).
Step 3
Match
Semantic retrieval finds candidate opportunities; an LLM reranks them and writes the rationale.
Step 4
Agents
Specialised agents suggest collaborators, propose proposal angles, and flag eligibility risks.
Step 5
Deliver
A weekly email plus this portal, where faculty give feedback that tunes future matches.

Also see the sign-up flow:Sign up Magic-link login