For research centers
The grant funded the science. Nobody budgeted the reporting.
AI implementation for NSF and DOE research centers and university research offices. The grant is the budget and the PI decides, so nothing waits on institutional procurement.
MRSEC, PREM, DMREF, EPIIC, NeuroNex, MIP, and DOE EFRC. We have run Division of Materials Research sites for close to a decade.
01
Who this is for
- Center directors and PIs
- Running an MRSEC, PREM, DMREF, EPIIC, NeuroNex, MIP or EFRC with a reporting and broader-impacts obligation that currently runs on somebody's evenings.
- Center coordinators
- You are almost certainly doing two people's jobs. The part we take is the chasing, the transcribing, and the version control, not the judgment.
- Research development offices
- NORDP-active administrators and ARIS member institutions who want AI inside the proposal and evaluation workflow rather than in a pilot that never ships.
02
What we build
- Annual reporting that assembles itself
- Center data in, drafted narrative sections out, review in the middle. A three-week scramble becomes a two-day read. The report is the deadline everyone plans around, so it is usually the right first build.
- Data collection that does its own chasing
- Structured requests, pre-filled with what is already known, sent to researchers who never log in to anything. Reminders on a schedule you set once rather than on a morning you remember.
- Outreach chatbots with real citations
- Retrieval over your actual corpus, answering with links back to the page, the paragraph, or the second of the video. Nothing asserted that a user cannot check in one click.
- Broader-impacts assessment
- Structured scoring built to survive a site visit and a reviewer who asks how the number was derived.
- Grant-advisor assistants
- Proposal support grounded in the solicitation, the program guidance, and the office's own funded proposals, so the advice is specific enough to act on at four in the afternoon.
- SimplyScholar, productized
- The agentic operations layer for a center, licensed annually and budgetable straight into the award. simplyscholar.com
03
How we engage
- Recon and proof, 6 to 8 weeks
- Fixed-fee proof of concept against one real workflow, with a written brief that fixes scope and withdrawal conditions before anyone opens an editor.
- Direct action, 3 to 6 months
- Implementation build, deployed into your environment, with your team trained on it before we leave.
- Standing capability
- SimplyScholar as a budgeted line inside the award, so the operating layer is funded for the life of the grant rather than re-argued every year.
04
Reference cases
-
Four weeks of annual reporting, down to a two-day review Write-up soon
The report now assembles itself from data the center already has.
NSF research center
In production -
Proposal help that knows the solicitation Write-up soon
Grounded in the program guidance and the office's own funded proposals.
University research office
In production
Which deadline hurts most?
Tell us which report, which review, or which workflow is eating the most time. We will tell you whether it is worth building for.