Gutcheck.AIExecution-Risk Bureau
Talk to us →
Built for accelerators, CDFIs, and workforce programs

A 0–1000 score built from six things, not one survey.

Gutcheck scores a founder's execution on six things — Volume, Consistency, Trend, Recency, Responsiveness, Reliability — pulled from the tools they already use, with their permission. No survey. No gut feel. A number, and the six reasons behind it, that your institution can show its own funders.

GUTCHECK.AI·EXECUTION RECORD
Execution Record · Doc 7F2A91CVerified 2026-08-20 · 14:22 UTC
Amara Osei
Founder Accelerator — Fall 2026
Verified Record
Execution Record
820
Confidence Grade: A
Strong Trend / Weak Recency
founder was accelerating, but has gone quiet recently
Illustrative record · this institution's name stands behind every record it references · full methodology on request
Method

Evidence, not opinion.

Five steps, end to end, no black box. Here's exactly how one consented data connection becomes one auditable Execution Record.

01

Connect

The founder authorizes read-only access to tools they already use — Slack, Gmail, GitHub, and more. No new survey, no manual entry, nothing extra to keep up with.

Sources
SlackLive
GmailLive
GitHubPending
02

Extract

On a schedule, Gutcheck pulls structured events from each connected tool — timestamps, counts, and metadata. Never message content, never read what was said, only that something happened.

Raw events
03

Normalize

Every event stream is scored against the same six observable features — for every founder, on every connected app, the same way.

A founder who's active but erratic scores differently than one who's quiet but reliable, and the record shows exactly which is true, not just a single number that hides the difference.

What's behind the number
01VolumeAny activity, at all
02ConsistencySteady or bursty
03TrendRising or fading
04RecencyHow fresh it is
05ResponsivenessSpeed of follow-through
06ReliabilityCommitments kept
04

Combine

The six features combine into one Execution Score, scaled 0–1000 — the same formula, every founder, every institution, no manual weighting, no thumb on the scale.

Execution Record
820/ 1000
Confidence Grade: A
05

Gate & grade

If too few features were actually observable, the record doesn't ship — a low score is real signal, a missing one means we don't know yet. Every record that does ship carries a Confidence Grade (A–E): not a judgment on the founder, but on how much independently-observed evidence backs the number. More sources, over a longer window, earn a higher grade — thin evidence still gets a score, just labeled honestly.

Confidence grade
A
B
C
D
E
Record issued
Above confidence minimum
Withheld
Below confidence minimum
Why now

AI broke the old signals.

Résumés, pitch decks, and applications no longer reliably distinguish execution from narrative. Institutions face rising accountability and bias scrutiny with no defensible evaluation infrastructure to meet it.

Status quo
  • Self-reported surveys and applications
  • Manual spreadsheet reconstruction at reporting time
  • No audit trail a funder can independently verify
  • Evaluation resets every cohort — no longitudinal history
With Gutcheck
  • Real activity founders permission us to see, not self-report
  • Always-current Execution Record, no manual reconstruction
  • Standardized, auditable — built to show a funder directly
  • Portable history that follows the founder across programs
Next

Bring Gutcheck to your cohort.

If you run a university entrepreneurship center, accelerator, or workforce program and need defensible evidence for how you evaluate the founders you serve — this is built for you.

Gutcheck.AI
Ratings infrastructure for institutional founder evaluation.