01
Role
Fydell and the employer define the work, judgment, communication, and environment the role actually requires.
From real work to a better interview.
Fydell does not grade a first answer. It observes how someone works, what happens when the situation changes, and whether the final decision holds up under questioning.
02
Work
The candidate enters a realistic job situation with resources, stakeholders, and an open work artifact.
03
Change
A deterministic material fact arrives. Fydell records what the candidate changes, preserves, and communicates.
04
Defense
Follow-up questions target the candidate’s actual decisions, contradictions, and remaining uncertainty.
05
Evidence
The work trail becomes claims with support, counterevidence, confidence, and explicit limits.
06
Interview
The employer gets the few candidates worth meeting and exactly what the next conversation should investigate.
Work observed
The candidate works in a realistic environment with files, people, assumptions, and an artifact that can change.
Evidence assembled
Actions and revisions stay connected to their sources, counterevidence, timestamps, and stated limits.
Decision verified
A reviewer checks the claim before the employer receives a brief and evidence-linked interview plan.
The strongest signal is what changes.
New material information forces the recommendation to move. The original assumption, revised artifact, reaction time, and remaining uncertainty stay visible together.
## Rollout sequence
10Launch all 1,200 seats in one production cutover at week six.
11Authentication review runs in parallel with production access.
10Weeks 1–6: sandbox tenant while the security review completes.
11Week 7: controlled 200-user cohort after security sign-off.
12Expand by business unit after the first support review.
## Open assumption
18Cohort size remains provisional until weekly active-user data is verified.
- assumption_addedweekly_active_users: unverified
- constraint_deliveredsecurity_review: six weeks, sequential
- artifact_revisedrollout_plan.md → revision 4
The brief is only as strong as the evidence beneath it.
Supporting events, counterevidence, oral defense, model and rubric versions, and human review travel with the claim.
Claim
The rollout plan was adapted to account for a six-week authentication security review, and the tradeoff was defended in the candidate’s own words.
Evidence lineage
Six-week timeline assumes no pre-launch security review.
Sandbox-first approach accommodates the six-week review.
Explained the tradeoff and why the constraint was retained.
Sponsor estimate assumed no extended review.
Supporting evidence
- Plan revised to sequence a sandbox-first approach.
Adjusts sequencing so production access follows the security review.
rollout_plan.md - Constraint received from the security team.
Security review is mandatory before production access.
security_brief.pdf
Counterevidence
- Sponsor estimate uses a shorter timeline.
Estimate assumes no extended security review.
sponsor_estimate.xlsx - Industry benchmark shows longer cycles.
Similar companies report six to eight week reviews.
benchmark_2026.pdf
Human review
K. PatelSolutions Engineering lead
ApprovedNotes
Tradeoffs explained clearly. Constraint retained and plan updated accordingly.
Reviewed 17:32
Open full evidenceArtifact
rollout_plan.mdline 14
Generated question
Your adoption assumption comes from the sponsor. How would you test it before using it to size the first production cohort?
Live transcript
- Candidate
I'd look for active-user data from teams similar to the sponsor and compare it to their license assignment.
- Interviewer
And if that data isn't available?
- Candidate
I'd size the first cohort smaller and treat the sponsor's number as directional, then validate with early usage.
rollout_plan.md · line 14
Competency
Commercial judgmentBring one open role. See the whole system work.
We calibrate the role, run the work, review the evidence, and return the people worth interviewing.