Build the architecture.
Match the benchmarks.
Show the reasoning.
A 16-step AI pipeline matches your internal roles to Mercer, WTW and Radford benchmarks — every match carrying a confidence score and a plain-English rationale. Job Architecture builds on that same validated work: principles, your level model, an AI-proposed straw model, the client deck and a full Global Job Catalogue.
GDPR compliant. Enterprise-grade security.
Built for compensation professionals,
by compensation professionals.
Designed by comp & rewards practitioners · Built for the work you actually do
From upload to export in five steps
No implementation project, no training data to prepare. Upload your files and the pipeline does the rest.
Set Up Project
Pick the survey vendor and the grading method — matrix read-across against your own level model, or a software-generated grade.
Your method, your levels.
Upload Jobs
Import your job list as CSV or XLSX. Column mapping proposes the match for every header — deterministically first, AI only for what the heuristics cannot settle — and you confirm it on one screen.
One screen, any headers.
Load The Survey
Add benchmarks from Mercer, WTW or Radford — mapped the same way — or reuse a survey catalogue you have already saved to your library. A file already in Plinto's template format needs no mapping at all.
Upload once, reuse all year.
Review Matches
Every row carries a confidence score, the grade the engine landed on, and a written rationale. Duplicate and consistency checks flag what needs you.
Every match shows its work.
Export & Validate
Download a five-sheet XLSX your stakeholders can read — every match, the auto-matched, what needs review, the abstentions, and a summary. Every match you validate is kept as your company's reference record.
Your decisions, on the record.
Set Up Project
Pick the survey vendor and the grading method — matrix read-across against your own level model, or a software-generated grade.
Your method, your levels.
Upload Jobs
Import your job list as CSV or XLSX. Column mapping proposes the match for every header — deterministically first, AI only for what the heuristics cannot settle — and you confirm it on one screen.
One screen, any headers.
Load The Survey
Add benchmarks from Mercer, WTW or Radford — mapped the same way — or reuse a survey catalogue you have already saved to your library. A file already in Plinto's template format needs no mapping at all.
Upload once, reuse all year.
Review Matches
Every row carries a confidence score, the grade the engine landed on, and a written rationale. Duplicate and consistency checks flag what needs you.
Every match shows its work.
Export & Validate
Download a five-sheet XLSX your stakeholders can read — every match, the auto-matched, what needs review, the abstentions, and a summary. Every match you validate is kept as your company's reference record.
Your decisions, on the record.
Now build the architecture,
not just the match
Job Architecture takes the projects you have already matched — across every survey vendor you use — and turns them into a structure your executive team can sign off. Multi-survey input is the whole point: one project per vendor by design, one architecture on top of all of them. Sources, principles, your level model, the straw-model loop, the client deck and the Global Job Catalogue are all live today.
A job is the deepest grouping, plus a level
Your layer names, your codes, your level model. Roles and positions sit below the job and stay out of the architecture — deliberately.
Job Architecture is available on the Max plan.
Choose your sources
Select any completed or validated matching projects as inputs. Evidence from every survey vendor you use lands in a single architecture instead of parallel spreadsheets.
Answer nine guiding principles
How many layers and what you call them. Detailed or streamlined. Code format and code logic. Catalogue depth. Your titling protocol. And your own level model — Plinto ships no default ladder, on purpose.
Read the straw model
Two proposed family-group trees — one detailed, one streamlined — each with grouping and estimated job counts. A gap panel names every survey grouping the model left unplaced, so nothing disappears quietly.
Make it yours
Rename, add, delete, merge, re-parent, rewrite a definition — or simply describe what is wrong and ask for another pass. Every draft is retained, so you can always show how the model got here.
Export the client deck
An 18-slide straw-model PowerPoint, filled deterministically from the approved tree — the deck you would otherwise spend a week assembling for the steering committee.
Generate the Global Job Catalogue
Freezes the approved straw model and generates the styled catalogue workbook, with a self-check report that blocks delivery on a structural failure rather than handing you a broken file, and every generated version kept.
Built for compensation teams that demand accuracy
Trusted infrastructure, transparent AI, and results you can actually defend in the room.
Signal extraction, discipline match, grading, consistency check and written rationale — each step recorded, each one inspectable.
The matching pipeline finishes a 140-job project in about five to six minutes. No manual spreadsheet mapping anywhere in it.
Every match ships with a confidence score, the grade the engine settled on, and the plain-English reason it got there.
Two plans, both built for real workloads
Every plan includes the full matching pipeline and both grading methods. What changes is how much you can run — and whether you get Job Architecture.
- 5 matching projects per month
- Up to 500 jobs per project
- 50 pipeline runs per month
- 5 saved survey catalogues
- 100 AI column-mapping calls per month
- Confidence scores, rationale and five-sheet XLSX export
- Validated matches kept as your reference record
- Everything in Pro
- 20 matching projects per month
- Up to 3,000 jobs per project
- 200 pipeline runs per month
- 25 saved survey catalogues
- 500 AI column-mapping calls per month
- Job Architecture — 5 architecture projects and 100 straw-model runs a month (deck exports included), plus 25 Global Job Catalogue generations
- Priority support
Monthly caps reset on the first of each calendar month. Saved catalogues and jobs-per-project are standing limits, not allowances — archiving a catalogue frees its slot.