Selected case study
Report 03 / Job architecture and compensation intelligence
Pilot-stage product with deep product, compliance-readiness, security, and pricing systems.
03
Job architecture and compensation intelligence
A Spain-first platform for role architecture, job evaluation, compensation decisions, pay transparency workflows, reviewable evidence, exports, and compliance readiness.
Compensation work needs traceable decisions, not just outputs.
The product starts from HR, legal, finance, and compensation teams that need explainable decisions, source maps, review workflows, and evidence they can defend.
Decision chain
A high-stakes domain where product copy must avoid overclaiming.
The system became a living compensation workspace.
The product spans puestos, job descriptions, factor valuation, employee mapping, imports/exports, compliance modules, evidence vaults, intelligence, pricing, and entitlement architecture.
Product modules
Real-codebase product leverage in a regulated, data-heavy domain.
Security and governance sit inside the product promise.
Hardening covered dependency risk, tenant employee references, CSV import limits, validation schemas, security headers, auth regressions, source governance, and privacy-aware reporting.
Governance checks
Responsible product work: do not turn sensitive data into casual automation.
The learning loop is evidence quality.
The strongest signal is the corpus of validation runs, manual review batches, pricing tests, source-policy decisions, compliance gap analysis, and roadmap discipline.
Evidence loop
Evidence-guided execution without pretending the product is legally complete.
Compensation decisions are high-stakes: teams need role architecture, pay evidence, regulatory interpretation, reviewer confidence, and privacy-aware workflows without turning sensitive data into casual automation.
A compensation-intelligence product with role catalogs, factor valuation, employee mapping, compliance views, source maps, evidence packs, pricing/entitlement logic, exports, and security hardening.
That I can convert regulation, methodology, data sensitivity, and evidence quality into product behavior while keeping human review and claim boundaries explicit.
01 / Builder signal
Gradimio shows how I work in a sensitive B2B domain where source grounding, auditability, privacy, human review, legal boundaries, pricing, and product architecture matter as much as the interface.
Gradimio is the clearest evidence-heavy B2B case. The product could not simply show compensation charts and call them intelligence. In this domain, the hard part is deciding what the chart means, what it must not claim, and what source or workflow supports the decision.
The work translated Spanish and European pay-transparency obligations into product objects, reviewer states, caveats, evidence packs, permissions, reports, and unresolved decisions. That separation matters because a useful internal metric is not automatically a legal conclusion.
The case shows product management before modification: understand the repository, map the domain, name the risks, define evidence, and only then turn selected parts into product behavior.
02 / Product evidence
Screenshots selected as product evidence, not decoration.
Curated live-product captures, cropped for readability so the product proof stays inspectable.
The page frames what the user is seeing before asking them to interpret the compensation data.
Charts, filters, legends, and tables remain connected so the product does not turn sensitive decisions into black-box output.
Compliance, organization context, and settings remain visible because the product sits inside a governance workflow.
The matrix gives HR and compensation teams a way to compare roles without hiding the criteria.
Compliance status, blockers, warnings, and action areas are structured for review instead of buried in data tables.
The detailed role screen exposes the valuation logic so sensitive compensation work stays explainable.
03 / Strategy choice cascade
This section is the product-management read: who the product is for, where it starts, how it can win, what capabilities matter, and how the work keeps learning.
Make job architecture and compensation decisions traceable, explainable, and compliance-ready.
Spain-first HR, compensation, legal, leadership, and consultant workflows.
Turn expert-heavy compensation work into factor evaluation, role architecture, exports, evidence, and intelligence.
Auth, Postgres/Prisma, employee data model, factor evaluations, audit history, exports, compliance modules.
Rebuild/parity planning, weekly product updates, pricing modules, validation scripts, security monitoring.
04 / Product map
Solo product builder across domain research, regulatory modeling, data architecture, UI, pricing, security, and validation workflows.
05 / Product-risk model
The risk map shows what had to be true before the product could be treated as more than a concept.
Requirements map to Spanish and EU pay-transparency obligations and related evidence workflows.
The product divides roles, data, compliance modules, reports, and reviewer workflows into distinct surfaces.
Docs define legal sources, obligations, remuneration concepts, equal-value groups, registers, artifacts, and audit events.
Three public plans and employee bands map product scope to company size and compliance needs.
Security hardening covers tenant references, CSV limits, schema validation, headers, auth regressions, and privacy controls.
The salary-intelligence work deliberately started with a small source-cited evidence pack rather than a broad automated benchmark promise. The point was to learn whether reviewer-facing evidence could be useful before making stronger claims.
That pattern repeats across the case: methodology versioning protects historical valuations, tenant-reference validation protects boundaries, and regulatory copy avoids collapsing distinct obligations into one vague compliance score.
06 / Product learning stories
Interview-grade product stories: what was ambiguous, what I chose, and what the product learned from that decision.
07 / Method bridge
Each flagship case shows the same pattern: use AI for leverage, keep judgment human, reduce product risk, and leave an operating loop behind the product.
The work depends on product briefs, real-codebase context, implementation review, validation checks, and careful claims rather than one-off prompting.
The case reduces the major product risks while making the trust surface explicit: value, usability, feasibility, viability, and launch confidence.
The next step is better evidence: user behavior, pilot feedback, operational checks, security posture, and roadmap decisions.
08 / Operating loop
09 / Trust layer
10 / What this shows