Generic AI screeners treat “shipped React for 5 years” and “followed a React tutorial” the same way. ShortlistTable distinguishes production stack from listed-skill scope, surfaces team size and on-call ownership, picks up GitHub and conference talks as evidence. AI screening with the depth technical recruiting actually needs.
| Candidate | Stack prod ≥ 3y | System design scope | On-call ownership | Priority | |
|---|---|---|---|---|---|
Alice Chen Ledger Platform · 7y | Kafka · Go · 5y | Tier-1 outage | Primary · 18mo | Call first | 1st |
Marco Silva Senior Cloud Arch · 12y | Stack varies | Multi-team | Secondary · 6y | Review | 2nd |
Nina Patel Indie · 4y | Tutorial-level | Solo apps | None | Hold | 3rd |
Technical recruiting fails when the screening criteria are written by someone who hasn’t built the system. Generic tools encourage generic columns — “strong CS background”, “experience with cloud”, “good engineer” — which produce noise that the engineering hiring manager ignores in favour of reading every resume themselves.
We give you a column system that supports the specificity technical roles actually need: stack maturity by years AND by scope, system-design experience by impact AND by team size, on-call ownership by primary vs secondary AND by duration. Written as plain-English questions, not as taxonomies.
Job titles alone are misleading at every senior+ level. We surface the title, the company stage at the time, and the team size if mentioned — so the recruiter judges “Senior” in context, not at face value.
Resume keyword scans treat both the same. Our extraction distinguishes production scope from listed-skill scope by reading the experience section, not the skills list.
Where the candidate’s public work is relevant (open source, conference talks, published papers), we surface it as evidence rather than letting it sit unread under “Links”.
Distinguishes production stack from tutorial / hobby use. A candidate who shipped React for 5 years is verdicted differently from one who lists React in a course.
Surfaces the title AND the company stage AND the team size, so the recruiter judges seniority in context — not at face value.
Reads beyond “designed system X” to ask: at what scale, with what team, owning what outcome. The verdict reflects the scope, not the verb.
Distinguishes primary vs secondary on-call, with duration. A candidate who has owned on-call for 18 months is verdicted differently from one who has been in rotation casually.
GitHub URL? We pick it up. Conference talk? Surfaced. Published paper? Cited as evidence. Public work becomes part of the per-cell evidence.
Write “2+ years production Elixir” as a column and the engine reads every resume against it. Not bound to a fixed taxonomy.
| Property | Keyword-based ATS | Generic AI ranker | Spreadsheet | ShortlistTable |
|---|---|---|---|---|
| Production vs tutorial stack | ✕Same weight | –Limited | ✕Manual | ✓Distinguishes scope |
| Title + scope + team size | ✕Title only | –Limited | ✕Manual | ✓All three surfaced |
| Obscure stacks supported | ✕Taxonomy-bound | –Depends on training | ✓Manual | ✓Plain-English |
| Public artifacts surfaced | ✕No | –Sometimes | ✕Manual | ✓URL + evidence |
| On-call distinction | ✕No | ✕Rare | ✕Manual | ✓Primary vs rotation |
Yes — the column logic is semantic, not a fixed taxonomy. Write “Has the candidate built a production system in Elixir for at least 2 years?” and the engine reads each resume against it the same way it would for Python.
If the resume includes a GitHub URL, we’ll detect it. Deep repo analysis isn’t in scope — for that you’d want a different class of tool — but presence of a public profile is captured as a column.
We surface the title, the company stage at the time the candidate worked there (when detectable), and the team size if mentioned — so the recruiter can judge “Senior” in context rather than at face value.
Yes. Write the column as a question: “Has the candidate owned a system serving > X traffic, with on-call ownership?” The engine looks for narrative evidence, not just keywords.
We don’t over-weight company brand. The columns can be written to focus on team size, scope, and on-call ownership — so a candidate from a private company who owned the right scope ranks correctly.
Yes. The column system is domain-agnostic — write the questions in the language of the role. Data engineering, ML engineering, security engineering have all been used by customers.
Try ShortlistTable on a 25-resume engineering pile. Free, no credit card, full column customisation.