Everything Capable can do, organised by what it's for.

Strength-first reports with every rating linked to specific evidence. Practitioner, candidate, and signed-off variants share the same source.

Cross-instrument profile per participant — interview, simulation, 360, cognitive, and personality results triangulated in one view.

Two finished assessments side-by-side. Rater drift is named, not hidden — the surface is built to make divergence legible.

Rate shared evidence against benchmark responses, surface dissent before it lands in real ratings, build your own accuracy.

The structured multi-assessor meeting that resolves dissent before sign-off — instruments sit side-by-side, evidence is named, the consensus is recorded.

Cohort drill-down, longitudinal deltas, benchmark ranges (research-anchored until consented cross-organisation data accrues), assessor performance, instrument coverage. Charts with date-range controls and a print-friendly export.

Heatmaps and gap views across teams and roles, with consent-aware aggregation and drill-through to the individuals you can see.

One query across every transcript passage and note you're accountable for — ranked, highlighted, grouped by person or skill, consent-aware throughout.

Gap and floor views against role profiles at organisation level — aggregate first, honest about thin evidence, drill-down bounded by consent.

Supply-drift projections over the workforce baseline — attrition, hiring, and development assumptions you set, with the assumptions named beside the outputs.

Sign-off, contest, correction, and reassessment queues with SLAs, reviewer routing, and a notifications inbox tied to each.

Three named pathways for a participant to push back — disagreement with a rating, factual fix to the record, fresh measurement.

Per-attribute disparate impact and counterfactual audit run live today; equalised odds and calibration join them as the outcome corpus accrues. Every metric is anchored to plain-language bands so it stays readable.

Drift, bias, and reliability indicators across every instrument in use. Surfaces the moment an instrument starts behaving differently from its norms.

Seven sector packs flex tone, distress thresholds, consent shape, audit-ready posture, and example phrasings. One choice per deployment.

Two shapes — quarterly assurance for boards and auditors, plus case-specific bundles for tribunals or regulators asking about one decision.

A hash-chain record of consequential actions — consent changes, role changes, programme transitions, sign-offs, key admin moves.

Every AI-derived artefact has a row — the input, the model family, the surface that produced it, the surface that consumed it.

Re-run an evidence bundle with a single attribute changed; surface the cases where the rating moves. Runs inside fairness monitoring on a routine cadence.

Every consent grant — interview tokens, magic-link grants, retention scope — in one auditable place, with revocation visible.

Combine any instruments in any order around one person or one cohort. Six sector-keyed templates ship in the gallery; you can author your own.

Portable credential the participant owns and shares. Signed, verifiable, translatable across SFIA, ESCO, and O*NET.

One cross-instrument view per participant. Interview, simulation, 360, cognitive, and personality findings sit alongside each other.

Every practitioner gets a calibration set seeded with their sector. Practice rating shared evidence; surface dissent before it hits production.

Every practitioner organisation arrives with a fictional sample organisation — invite a fictional participant, draft a report, sign off, respond to an appeal, all without touching real data.

Seven sector packs flex tone, palette, distress thresholds, consent shape, and audit posture — government, corporate, community, healthcare, education, finance, plus an Indigenous-led variant.

Five locales today — en-AU, en-GB, en-US, zh-CN, vi-VN — with the help centre authored in MDX so non-engineers can translate per locale.