Build · Create · Ops & Automation
Bias-guarded interview debrief packet
A structured debrief record that captures independent pre-discussion verdicts, a bias-check, and a consolidated recommendation.
You receive: A JSON object: { role_title, candidate_label, bias_check_prompt_field, independent_verdicts[] (each {interviewer_label, lean enum: hire|no-hire|unsure, evidence_field, recorded_before_discussion bool}), consolidated {decision enum: advance|reject|hold, rationale_field, evidence_refs[]}, backup_candidate_field }.
Part of Lead Team
Example
A sample of what this play produces. Your result is generated for your inputs.
- Role title
- Founding Engineer, Judge and Fulfillment Infrastructure
- Candidate label
- Candidate A (fictional example candidate)
- Bias check prompt field
- Before anyone speaks: where could bias show up in this loop? Candidate A pattern-matches to a big-lab background, which risks halo bias on the systems round. Two interviewers saw the take-home score before their own sessions, which risks anchoring. The role is remote-first, so discount any 'energy in the room' impressions. Name your own likely bias out loud before the discussion opens.
- Backup candidate field
- Candidate B (fictional): weaker on sandbox isolation depth but stronger buyer empathy; hold warm and re-engage within two weeks if Candidate A's work-sample week does not confirm E1.
Independent verdicts
| Interviewer label | Lean | Evidence field | Recorded before discussion |
|---|---|---|---|
| Interviewer 1 (Dan Schmitz, founder) | hire | Work sample: extended the sandboxed judge harness with a process-isolated timeout path and wrote failing tests first (E1). Explained the acceptance-check tradeoffs in plain language a buyer could follow (E2). Directly relevant to relieving the concierge bottleneck (1 founder-operator, about 30 plays/week today). | true |
| Interviewer 2 (advisor, systems round, fictional) | hire | Whiteboarded the play-run settlement flow and correctly identified the validator floor vs quality distinction without prompting (E3). Proposed a queueing design for judge workers that matched our real Vercel constraints (E4). | true |
| Interviewer 3 (advisor, product round, fictional) | unsure | Strong on infrastructure but gave thin answers on buyer-facing tradeoffs: could not articulate who the target buyer is beyond 'developers' (E5). At pre-seed the founding engineer will talk to buyers weekly, so this matters. | true |
Consolidated
- Decision
- advance
- Rationale field
- Two independent hire leans on the core competency (judge and fulfillment infrastructure, the capacity bottleneck for a pre-seed team of one) outweigh the product-round unsure, which is coachable and partially offset by the founder covering buyer contact. The unsure verdict was recorded before discussion and its evidence (E5) is logged as an onboarding focus, not dismissed. Advance to a paid work-sample week scoped to the judge worker queue.
- Evidence refs
- E1
- E2
- E3
- E4
- E5
Opens soon
Cost50 credits
ProtectionSpent when the run starts · returned only if STUD is at fault
STUD opens runs for everyone soon, on every play at once.
Get early access to STUD the day it goes live.