GrantScout — Tetelestai

The cost of grant-seeking

116 hrs

Average time a principal investigator spends writing a single grant proposal, per a survey of 195 astronomers and psychologists applying to NIH, NSF, and NASA. von Hippel & von Hippel, PLOS ONE, 2015

34 days

Average researcher time spent preparing a single proposal, per an observational study of NHMRC (Australia) grant applicants — with no measurable link between time invested and odds of success. Barnett et al., BMJ Open, 2015

~20–25%

Typical success rate for competitive grant programs cited across the literature — meaning the majority of proposal-writing hours go toward applications that are never funded. Multiple sources, incl. NHMRC and NSF program data

45%

Estimated share of a principal investigator's time spent on administrative activity tied to applying for and managing funded projects, rather than active research. Cited in Vaesen & Katzav, arXiv, 2024

How it works

A cost-conscious pipeline — cheap filters first, expensive reasoning only where it's earned.

Continuous discovery. Grants.gov, EU Funding & Tenders, foundation pages (Wellcome, EMBO, HFSP, Fondazione Cariplo, AIRC, Telethon Italy, and whatever's relevant to your own region), RSS feeds, and an AI-driven web search run on a rolling schedule — not just when someone remembers to check.
Free filtering before any AI runs. Expired deadlines, hard geography mismatches, and zero keyword overlap with your lab's profile get dropped by rule-based filtering — no cost incurred on grants that were never in play.
Two-pass AI review. A cheap relevance check clears out what's obviously irrelevant; only genuine candidates reach the expensive model that extracts deadline, effort level, eligibility, funding amount, and — critically — whether your PI can actually be the named applicant, not just technically host someone else's fellowship.
Deep scoring on the real contenders. The top candidates get scored across six weighted dimensions — topic relevance, eligibility, geography, career-stage fit, and methodology — then ranked and geography-tiered to your lab's actual location, so a lower-odds long shot doesn't outrank a stronger local fit.
It learns from you. Every grant you mark relevant or not feeds a feedback model that nudges future rankings toward what your lab actually pursues — cold-start-safe, so it does nothing until it has enough signal to trust.

Why not Pivot-RP, GrantForward, or Instrumentl

The existing options were built for a different buyer than a single lab.

Pivot-RP (Clarivate) and GrantForward are the two databases most researchers actually encounter — usually through an institutional subscription their university already pays for. Instrumentl and similar AI-matching tools serve mostly US nonprofits. GrantScout isn't built around one country or one funding system — it finds and ranks sources against wherever your lab actually is, whether that's Bologna, Alabama, or Dhaka.

Capability	GrantScout	Pivot-RP / GrantForward	Instrumentl & general AI tools
Access	Direct, single-lab	Institutional subscription only — priced by university size	Direct, but built for US nonprofits
Adapts to your lab's own location	Yes — any country, any lab	Thin outside US/UK	Thin outside US
Flags the PI-eligibility trap	Yes	No — read the fine print yourself	No
AI cost applied selectively	Yes — free filters run first	Editorially curated, not per-listing AI	Runs broadly across listings
Learns from your lab's own accept/reject history	Yes, lab-specific	Institution-wide profile matching	Keyword/eligibility matching

Grant search takes weeks of PI time. Most of it is spent on proposals that were never a fit.

A cost-conscious pipeline — cheap filters first, expensive reasoning only where it's earned.

The existing options were built for a different buyer than a single lab.

Currently onboarding a limited number of labs.