Legacy SaaS dossier · greenhouse

Greenhouse

CategoryATS (recruiting)Greenhouse Expert + Sourcing add-onSite ↗

A candidate pipeline + scorecards + an audit log, sold as a per-employee tax. The audit log is the only piece you can't trivially replace.

01 · Data

Data the SaaS sits on

LinkedIn (sourcing)API

Interviewer calendarsYours

Job boardYours

Payroll / HRISAPI

Internal Build Score

0.56

Feasibility of a useful internal substitute for the actually-used 20% of this SaaS, built with Claude and your own data layer.

IBS = 0.25·U + 0.25·C + 0.20·D + 0.15·I + 0.15·Mthis record · 56%

U
User adoption gap
weight 0.250.55
How much licensed surface area sits idle.
C
Core simplicity
weight 0.250.60
How small the actually-used 20% is when re-implemented honestly.
D
Data portability
weight 0.200.50
How clean the exit is, exports, formats, audit lock.
I
Integration burden (inverse)
weight 0.150.55
How few external systems must keep working through the SaaS.
M
Moat erosion
weight 0.150.60
How much the original differentiator has been commoditised by LLMs + cheap infra.

Editorial heuristic. Not investment advice, not a procurement recommendation.

03 · Feature usage

Commonly used vs. rarely touched

Commonly used

Candidate pipeline with stages
Interview scorecards per role
Scheduling across multiple interviewers
Source-of-truth applicant record
EEOC / audit log

Rarely touched

CRM module for candidate nurture
Sourcing + Marketing as separate paid pillars
Approvals workflow with custom forms
Onboarding handoff layered on top
Reporting suite with a non-SQL query builder
AI screening, a Claude call billed per req
Greenhouse Expert / Pro / Essential ladder

Build it yourself→

Internal build

Build it yourself

01 · Stack & flow

Postgres

Next.js

Claude API

Calendar API

S3 (CVs)

Internal build map

Data in

Candidates table

Interviewer calendars

CV blobs (S3)

Agent layer

Scorecard synthesiser

JD writer

Logic

schedule loop

collect scorecards

append audit log

EEOC reporting

Outputs

Pipeline board

Debrief read

Audit log row

Build time: 9 weeks
Run cost: $5/seat·mo
vs. SaaS: $220K / yr

Build outline

Candidates + stages + interviews schema
Pipeline board + per-role scorecards
Interview scheduler (free-busy + round-robin)
Append-only audit log w/ row-level immutability
Claude scorecard synthesis after every loop

02 · Claude prompt

First, map what you actually use (optional)

Step 1 · usage audit prompt

// Greenhouse usage audit, scope the rebuild Input: the pipeline stage config with per-stage volume, the scorecard templates with fill rates, the configured approval flows, and the report catalog. ## Do 1. Keep stages candidates actually move through; collapse dead ones. 2. Keep scorecard attributes filled on > 20% of loops. 3. Mark the EEOC / audit-log requirements as load-bearing, they must be reproduced exactly, not trimmed. ## Output JSON: { keep_stages[], keep_scorecard_fields[], compliance_must_keep[], drop[] }.

Step 2 · rebuild the core

// Scorecard synthesiser, internal recruiting tool You receive (a) a candidate's full interview loop (each interviewer's structured scorecard + written notes), (b) the role's scorecard rubric, and (c) the hiring manager's stated must-haves. ## What you must do 1. Produce a debrief synthesis: - Per-rubric-area: average score + the supporting quote. - Bench strengths + risks, quoting interviewers, never paraphrasing. - Single recommendation: "hire", "no-hire", "down-level", or "need-another-loop", with one sentence of reasoning. 2. Flag any rubric area where interviewers disagreed by ≥ 2 points in `disagreements[]`. ## What you do not do - Do not name interviewers in the synthesis ("Interviewer A" is fine for debrief-room neutrality). - Do not infer protected characteristics; do not comment on culture fit unless the rubric explicitly contains a culture column. - Do not generate the offer letter, that's a separate flow. ## Output JSON: { per_area[], strengths[], risks[], recommendation, disagreements[] }.

03 · Result

Synthesise loop for Senior PM candidate, 5 interviewers

loop:2025-08#cand:1487

hire @ L5 · strengths: 3 · risks: 1 (depth on growth metrics) · disagreement: collaboration (2→4)

← Legacy SaaS index AI wrappers index →