Candidate scoring blueprint
Four scores, one verdict, and the reason behind each.
Enter a candidate and a target field, and get a fit score with the reasoning behind it. A personal project, inspired by production talent-intelligence work and built from scratch with its own logic and entirely synthetic data.
V1 · Oct 2026
The rubric
Four things every candidate is scored on, and what each band means.
| Dimension | Range |
|---|---|
| Technical Depth | 45 to 98 |
| Public Profile Strength | 45 to 98 |
| Leadership Signal | 45 to 98 |
| Domain Relevance | 45 to 98 |
Reasons shown for each band
| Band | Score | Reasons it can show |
|---|---|---|
| High | 78 or more | Strong, consistent signal across public sources. Well above the threshold for this dimension. Clear evidence of depth in this area. |
| Mid | 60 to 77 | Moderate signal — present but not standout. Meets baseline expectations for this dimension. Some evidence, could be stronger with more data. |
| Low | below 60 | Limited public signal for this dimension. Below the typical threshold — worth a closer look. Sparse evidence available for this criterion. |
How a score is made
From the button press to the verdict.
Swipe sideways to see the whole flowchart
Read this flow as a list
- Run Scoring pressed. A candidate name and a target domain
- Name entered?. The name is required
- Short loading sequence. Parsing input, running the sample rubric, computing the fit score
- Nothing runs. Focus returns to the name field
- Seed from name and domain. The name is lowercased and trimmed, joined to the domain and hashed
- Score four dimensions. Each from a fixed formula, between 45 and 98
- Pick a reason for each. High 78 or more, mid 60 to 77, low below 60
- Average into the fit score. The mean of the four scores, rounded
- 78 or more?. Strong Fit
- Strong Fit. Shown in green
- 60 or more?. Consider
- Consider. Shown in amber
- Not Yet. Shown in red
The screens
Three views, one top bar.
| Screen | Shows | You can |
|---|---|---|
| Dashboard | Candidates scored, average fit score, the four rubric dimensions and the five latest scores | Jump straight to scoring |
| Score a Candidate | A form with candidate name, target domain (5 choices) and an optional LinkedIn URL | Run scoring and read the four scores, the overall fit score and the verdict |
| Pipeline | A table of sample candidates with domain, score and status | Filter by domain or status, sort by score |
The sample pipeline
12 made-up candidates, scored by the same formula.
| Candidate | Domain | Score | Status |
|---|---|---|---|
| Jordan Blake | Developer Tools | 82 | Strong Fit |
| Taylor Novak | Fintech | 80 | Strong Fit |
| Priya Nair | Fintech | 76 | Consider |
| Morgan Diaz | AI / Machine Learning | 74 | Consider |
| Drew Halvorsen | Cloud Infrastructure | 71 | Consider |
| Alex Rivera | AI / Machine Learning | 70 | Consider |
| Robin Castellano | AI / Machine Learning | 66 | Consider |
| Jamie Kowalski | Design / Product | 65 | Consider |
| Casey Lindqvist | Cloud Infrastructure | 63 | Consider |
| Avery Lindgren | Fintech | 63 | Consider |
| Reese Andersson | Developer Tools | 62 | Consider |
| Sam Okafor | Design / Product | 55 | Not Yet |
What is real and what is not
The line between the idea and the sample.
| Part | In this demo |
|---|---|
| Candidate profiles | None are read, fetched or stored |
| Scores and reasons | Generated from the name and domain by a sample formula |
| LinkedIn URL field | Optional, and never fetched |
| Names in the pipeline | Made up |
| Network calls | None |
Want something like this?
Tell me about the process you want off your plate. The brief takes about three minutes, and the project is priced before I start.
Start a BriefScoutOS is a personal, independently built demo. It uses synthetic data only, and real records and anything confidential are left out.