Corsable Answers

A competency framework is the structured definition of what "good" looks like for a role, organised as a hierarchy of levels, departments, categories, topics, and rubric-backed requirements. Each requirement spells out an observable, scorable standard, so the same framework drives knowledge gates, live competency exams, and audit-ready evidence against one consistent rubric. It is the backbone every other part of Corsable references when verifying and recording skill.

Choose the "Start blank" path when creating a framework, set the name, domain, and primary or secondary language, then author the tree yourself in the editor. You add levels, then departments, categories, topics, and individual requirements as nested nodes, with inline add, edit, and delete on every node. You can also ask the AI to generate or regenerate any single node, so building by hand does not mean writing every line unaided.

Yes. The "Upload PDF" path takes an existing competency manual or SOP, runs OCR on scanned pages, detects the heading hierarchy, and synthesises a draft framework tree. It then opens a three-pane reviewer where you accept, regenerate, or edit each node before publishing, so nothing goes live unreviewed. PDF import is available on the relevant org plans; see Request access.

A framework nests five node kinds: levels (seniority bands such as junior to lead), departments, categories, topics, and requirements. Requirements are the leaf nodes and carry the rubric, the concrete, scorable standard a person must meet. Higher levels can build on lower ones, so progression is explicit and a learner advances by satisfying the requirements at each step rather than by tenure.

Every requirement in a framework is tied to a rubric that defines exactly what evidence and standard count as meeting it. That rubric is what knowledge gates and live AI competency exams score against, so a pass reflects a defined bar rather than a grader's impression. Because the rubric is fixed and recorded, the resulting scores are consistent across people and defensible in an audit.

Frameworks are built from reusable building blocks, levels, categories, topics, and requirements, that you can author once and adapt. Corsable also ships domain-tuned starter templates (for example Retail Banking, Corporate Sales, and Customer Support) that give you a populated tree of levels and topics to customise rather than starting empty. You restructure the skeleton freely after seeding, so a template or an imported framework becomes the basis for variants.

No. Corsable is domain-agnostic: the level, category, topic, and requirement structure fits any role in any sector, from retail banking tellers and customer support tiers to clinical, technical, or operational roles. The built-in templates cover a few common domains as starting points, but a blank framework or a PDF import lets you model any competency taxonomy. Frameworks can also be authored bilingually across 12 supported languages, including Arabic and Urdu.

It depends on the path you choose. Importing from a PDF produces a reviewable draft tree shortly after upload, then your time is mainly spent in the three-pane reviewer accepting or refining nodes. Starting from a built-in template gives you a populated structure to adjust immediately, while a fully custom blank framework takes longer because you author every level and requirement, though per-node AI generation speeds that up.

Yes. The framework tree editor lets you expand any node and add, rename, move, or delete levels, departments, categories, topics, and requirements at any time. You can also regenerate an individual node with the AI when you want to refine its wording or rubric. The initial seed, whether blank, template, or PDF import, is just a starting skeleton you are expected to reshape.

Yes. When you create a framework you set a primary language and an optional secondary language for a bilingual framework, with names captured per language. This supports right-to-left languages such as Arabic and Urdu alongside Corsable's 12 supported languages, so a single framework can serve a multilingual workforce. The same framework then drives gates, exams, and evidence in those languages.

It is an AI-assisted voice exam where the learner works through a realistic, role-specific scenario rather than answering multiple-choice questions. The AI examiner conducts the session in real time, probes the learner's reasoning, and scores their responses against the competency framework's rubric. It is the applied step that follows the knowledge gate, designed to verify that someone can actually do the work, not just recall facts.

Each exam is set in a realistic scenario tied to the role and competency level being assessed, so the learner has to apply judgement out loud rather than recite definitions. The AI examiner can follow up on vague answers, ask why a decision was made, and adapt to how the candidate reasons through the situation. Scoring is mapped to the rubric-backed requirements in your competency framework, so a pass reflects demonstrated capability against defined criteria.

Live exams are proctored with anti-cheat measures built into the session to keep results trustworthy. Because the exam is a real-time spoken conversation against an adaptive scenario, answers cannot simply be looked up or pasted, and the examiner can probe inconsistencies on the spot. The full session is captured as part of the tamper-evident evidence record for later review.

Responses are assessed against the rubric-backed requirements defined for the relevant competency, category, and level in your framework, rather than against a generic answer key. The result ties each score back to specific criteria, so a pass or fail is explainable in terms of what was and was not demonstrated. Scores are stored alongside the transcript and rubric in the audit-ready evidence record.

Yes. Every live exam produces a tamper-evident record including the score, the full transcript, the rubric used, and timestamps, all of which a human reviewer can examine. This keeps a person in the loop for sign-off, appeals, or spot checks, with the AI handling the conduct and first-pass scoring. The records can be exported as part of a compliance pack.

Each result is a tamper-evident record that pairs the score with the transcript, the exact rubric applied, and timestamps, so the basis for every decision is reconstructable after the fact. Combined with proctoring and anti-cheat during the session, this gives a defensible chain of evidence rather than a bare pass/fail flag. Records export into compliance packs for auditors. See /solutions/compliance.

Live exams are available in 12 languages, including Arabic and Urdu with full right-to-left support. The voice exam runs in the learner's chosen language so candidates are assessed on competency rather than on working in a second language. Corsable is domain-agnostic, so this applies across whatever roles and frameworks you assess.

Yes, that is the intended flow. Knowledge gates are adaptive multiple-choice gates that verify fundamentals first, with configurable pass thresholds and randomised question banks. Only once the gate is passed does progression open to the live AI competency exam, so the applied assessment is reserved for candidates who have already proven the basics.

Exams are tied to the competency framework, which structures a role into levels, categories, and rubric-backed requirements. The scenario and the criteria the candidate is scored against come from the level and competencies being assessed, so the same role can be examined differently at different seniority levels. You can build the framework from scratch or import one from a PDF.

Yes. The evidence captured during an exam, including scores, transcripts, the rubric, and timestamps, can be exported as a compliance pack for auditors, regulators, or internal records. Because the records are tamper-evident, the export reflects exactly what happened in the session. Corsable is GDPR ready and SOC 2 is in progress.

Every topic in a Corsable training program carries its own private AI tutor that a learner can open while they study. It is grounded in the exact item on screen, so it explains, clarifies, and quizzes against that specific material rather than answering from general knowledge. Learners reach it without leaving the module player, and conversations stay scoped to what they are currently working through.

The tutor is grounded in the precise item the learner has open, not the whole course or the open web. That includes the flashcards on the card deck, the text of an uploaded document or PDF, and even the narration of a video. Because the grounding is item-level, answers stay on-topic and reflect your own materials rather than a generic model response.

Yes. When a learner opens a video topic, the tutor is grounded in that video's narration, so it can answer questions about what was just said and reinforce the explanation. This works for AI presenter videos and for your own uploaded videos that carry narration. The learner can ask follow-up questions in the same session instead of rewatching to find a point they missed.

Corsable can generate flashcards, quizzes, mind maps, and slides for a topic from your source material. These AI-generated items sit alongside your own uploads in the same module, so a topic can mix generated study aids with your existing video, slides, or PDFs. You choose which generated items to keep and arrange them as steps within the module.

Yes. Flashcards, quizzes, mind maps, and slides are produced from the material you provide rather than from a generic question bank, so learners study what is actually in your content. The same grounding applies to the per-topic AI tutor, which references your uploaded items directly. This keeps both the generated study aids and the tutor's answers tied to your source of truth.

Yes. You can upload a PDF as a source, and Corsable reads it, including scanned pages, using vision OCR so the text becomes usable content. From there you can keep the PDF as a topic item and generate flashcards, quizzes, mind maps, or slides from it. PDF import is also how you can build a competency framework from an existing document.

Yes. A training program's modules and topics can mix AI-generated lessons with your own video, slides, and PDFs in any combination. Your uploaded items are first-class topic content, and the per-topic AI tutor is grounded in them just as it is in generated material. You can also set required steps and unlocks so learners move through your content in order.

Yes. Corsable supports 12 languages, including Arabic and Urdu with full right-to-left layout, alongside English, French, Spanish, German, Portuguese, Italian, Dutch, Turkish, Indonesian, and Malay. The platform is domain-agnostic, so the tutor and generated content are not tied to any single subject area. AI presenter videos are also available across the same 12 languages.

Yes. The tutor opens as a private session for the individual learner on the item they have in front of them. It is intended as a one-to-one study aid rather than a shared or broadcast channel. Conversations are scoped to the topic the learner is working through at the time.

AI presenter videos are narrated, subtitled explainer videos featuring a consistent on-camera presenter whose appearance is gender-matched to the voice, available in 12 languages. Slides, flashcards, quizzes, and mind maps are study items generated from your material, whereas a presenter video is a narrated lesson you can place as a topic step. You can also upload your own video in place of a generated presenter video. The tutor is grounded in the narration of whichever video the learner opens.

The tutor is deliberately scoped to the exact item the learner has open, so its answers reflect that flashcard deck, document, or video narration. This keeps explanations grounded in your material rather than drifting into unrelated general knowledge. If a learner needs help on a different topic, they open the tutor on that topic, where it grounds in the relevant content.

Every competency exam produces a structured record rather than a pass/fail flag: the scores against each rubric requirement, the supporting evidence quote the AI used, the full exam transcript, the rubric version applied, and the timestamps for each step. These records are kept as tamper-evident history, so an assessment decision can be reconstructed and defended after the fact. The aim is that a competency claim is backed by reviewable proof, not just a database tick.

A completed exam session stores the full transcript, the rubric state at the time of scoring, a score for every rubric requirement with the evidence quote that justified it, and the captured audio. Proctoring and identity signals from the session are retained alongside it. Together these let a reviewer or auditor see what was asked, what the learner said, and why each requirement was scored the way it was.

Audit and assessment records are written as append-only history: each audit event records the action, the actor, the affected resource, the originating IP and the timestamp, and is not edited in place. Scores, transcripts and rubric versions are preserved as captured rather than overwritten, so later changes are added as new entries rather than silent edits. This gives a traceable chain you can show a reviewer.

The audit log is a read-only, paginated record of administrative and system actions: for example who created or changed a framework, who assigned an exam, and the result. Each entry shows the action, the actor, any target, a timestamp, and supporting detail. It is designed for review rather than editing, so admins can see the history of who did what and when.

Assessment evidence covering scores, the rubric used, the transcript and timestamps is held in a structured form that can be exported for an individual or a cohort to hand to a regulator, an auditor or an internal quality team. Because the rubric version and evidence quotes travel with the record, the exported pack is self-explanatory rather than a bare score sheet. If you need a specific export format for a particular regulator, see our compliance solutions.

Yes, that is the core design. From a stated competency you can drill to the exam that verified it, then to the individual rubric requirements, the score on each, and the exact transcript passage cited as evidence. The rubric version and timestamps are attached, so you can show not just that someone was assessed but precisely what they demonstrated and when.

A completion record shows that someone finished a module or training program: they watched the videos, passed the quizzes and reached the end. Evidence of competence comes from a rubric-scored exam that verifies they can actually perform the role-relevant skills, with the transcript and per-requirement scores to prove it. Completion answers "did they do the training"; competence evidence answers "can they do the job", and Corsable keeps both distinct.

The evidence is built to be shown to a third party: it pairs each score with the rubric it was measured against, the supporting transcript quote, and a timestamp, so an outside reviewer can follow the reasoning without access to the live system. Corsable is GDPR ready and SOC 2 is in progress. Exactly what an auditor will accept depends on your sector, so review the exported pack against your own regulatory requirements.

Assessment records, including transcripts, scores, rubric versions and the related audit entries, are retained so that historical competency claims remain verifiable rather than expiring with a learner's session. Captured exam audio is stored in object storage as part of the review evidence. If you have a defined retention or deletion policy for your organisation, raise it with us so retention can be aligned to your obligations.

It includes both. Each requirement carries its numeric score plus the rubric dimension it was measured on and an evidence quote drawn from what the learner actually said, in addition to the topic-level and aggregate scores. This means a reviewer can see why a score was given, not only what it was, which is what makes the record defensible.

Yes. The live exam stores the full text transcript together with the captured audio chunks, and the audio is kept so the original capture order can be reconstructed independently of the text. A reviewer can therefore read the transcript and, where needed, listen back to the source audio. Both are held as review evidence for the session.

Yes. Each requirement has an AI-produced score and a separate final score, so a reviewer can confirm or adjust the outcome, and there is a review decision recorded on the session. Any such action is itself written to the audit history, so the human decision and its author are part of the evidence trail rather than replacing the original.

Yes. Individual learners can use Corsable for free, including taking training programs, working through topics with the AI tutor, and earning certificates on completion. Paid tiers apply only to organisations that need to build competency frameworks, run gates and live exams, and manage members at scale. See access for the full breakdown.

Organisation pricing is per-organisation, not per-seat, across three tiers: L&D Starter at $299/mo ($2,990/yr) for 25 learners and 1 framework, L&D Growth at $999/mo ($9,990/yr) for 250 learners, 25 admins and 5 frameworks, and L&D Enterprise at $2,999/mo ($29,990/yr) for 2,500 learners, 250 admins and unlimited frameworks. Annual billing is discounted against the monthly rate. Live prices are shown on access.

The tiers differ mainly in capacity: learner seats (25, 250, 2,500), the number of admins, and how many competency frameworks you can run (1, 5, and unlimited respectively). Growth is the common choice for teams certifying several roles, while Enterprise adds the highest limits plus provisioning features such as SSO/SAML and SCIM. All tiers include knowledge gates, live AI exams, training programs and audit-ready evidence.

Yes. Organisations can start a 14-day free trial to build their first framework, run a knowledge gate, and generate a sealed evidence record before committing to a paid tier. No long-term commitment is required to evaluate the platform during the trial.

Individual learners can sign up and start for free immediately. Organisations evaluating Corsable for a team can begin the 14-day trial or book a guided demo to walk through frameworks, gates and live exams with their own roles in mind. Use the Request access to begin or request a demo.

Corsable is GDPR ready and is currently pursuing SOC 2, which is in progress. The platform is built around audit-ready evidence: every gate result, live exam and assessor decision is captured as a tamper-evident, hashed record so any change is detectable. Details are outlined on the compliance page.

Corsable is GDPR ready, and learner and assessment data, covering scores, transcripts, rubrics and timestamps, is stored as part of each individual's competency record under your organisation's control. Evidence is exportable on demand into compliance packs, and records are hashed so tampering is detectable. See the privacy page for how data is processed and retained.

SOC 2 is in progress rather than complete, and Corsable does not claim certification it has not yet achieved. The platform is GDPR ready today, and its evidence model of tamper-evident, hashed records with full transcripts and rubrics is designed to support audit and compliance review in the meantime.

Corsable's interface supports 12 languages: English, Arabic, French, Spanish, German, Portuguese, Italian, Dutch, Turkish, Indonesian, Malay and Urdu. AI presenter videos can also be generated with narration and subtitles in these languages, with the on-camera presenter gender-matched to the chosen voice.

Yes. Arabic and Urdu are fully supported with right-to-left (RTL) layout across the interface, so menus, content and learner-facing screens render correctly. Corsable is domain-agnostic, so frameworks, programs and exams work the same way regardless of the language you author or learn in.

At the Enterprise tier, Corsable supports SSO/SAML for single sign-on and SCIM for automated user provisioning, so members can be added, mapped to job titles and seniority, and deprovisioned through your existing identity system. This keeps organisation membership in sync without manual account management. Speak to the team via a demo to confirm fit for your provider.

No. Corsable is domain-agnostic competency-based L&D, so the same framework, gate, live-exam and training-program tools apply to any role or industry. You define the levels, categories and rubrics that matter for your roles, and the platform works the same whether you are certifying technical, clinical, commercial or operational competencies.