A concept brief for the next product in the ReadyEngine platform. Phase 1 (ReadyRating) gets the licensee through the regulator's gate. Phase 2 builds and measures the competence the gate was a proxy for — what the regulator's stated goal actually wants, what the firm actually needs, and what the licensee actually requires to do the job well in an AI-mediated workplace. The same architecture serves regulated CE, safety and certification training, and competency-based development in any industry — deployable as a WKT-published product or as a licensable diagnostic platform that other organizations bring their own content into.
Prepared by Emma Plumb. ReadySignal is used throughout as the working name for the proposed Phase 2 product; ReadyPractice is the alternative; naming decision presented at the end. This brief stands alongside the ReadyEngine Strategic Brief (May 2026) as an exploration of where the platform could go next.
Three distinct failure modes of professional learning are converging in 2026, and the current generation of learning platforms is not designed to address any of them well.
The dominant design pattern in workplace training is what cognitive science calls expository content delivery: the learner sits through a video, reads a module, completes a quiz, and gets a completion record. The platform measures consumption (modules completed, hours logged, quiz scores) and assumes that consumption equals learning.
This design is comfortable to build, comfortable to consume, and largely ineffective for durable competence. The learner ends up with what cognitive science calls a pointer — they know where the information was and can recognize it when re-cued, but they have no schema they can retrieve under pressure or apply to a novel situation.1 Within weeks of the training event, retention has decayed substantially. Within months, almost nothing transfers to actual practice. This is the well-documented decay curve every L&D leader privately knows about and publicly works around with annual refresher courses.
This failure mode has been a known problem for decades. What is new is how badly it interacts with the other two.
In regulated industries — financial services, real estate, healthcare adjuncts, certain trades — and in mandatory safety and certification training, the L&D market has settled into an even shallower equilibrium. The regulator requires hours; the provider sells hours; the licensee consumes hours; the firm logs hours. Whether anyone has actually learned anything is not part of the transaction.
The unique challenge in this market is the structural pressure on duration. In ticket training (safety certifications, mandatory compliance refreshers, prerequisite courses), the worker wants to be in and out — they have a job to do and a few hours to spare because someone required them to. Multi-month engagement is the wrong product shape for that buyer. The platform has to deliver measurable competence development inside a constrained engagement window, not outside it.
The result is that regulated CE and ticket training have become almost entirely about the artifact — the credit, the certificate, the audit record — not about the development. The gap between completion and capability is widely understood inside the industry — by firms, by senior practitioners, by L&D leaders. Whether regulators themselves will eventually move toward outcome-based models is, at this point, a hypothesis worth testing rather than a confirmed direction of travel.
The third failure mode is new, accelerating, and the reason any of this matters now rather than five years from now. AI tools are entering professional workflows at scale — every junior advisor, every newly licensed broker, every fresh-from-orientation real-estate professional is now reaching for an AI to help them think.
The cognitive-science evidence on what this does to learning is already converging.3 When a learner has built durable schema before they reach for AI, the AI amplifies their thinking — they can evaluate its outputs, refine its prompts, catch its errors. When a learner reaches for AI before their schema has built, the AI substitutes for the thinking they never built — they mistake the AI's fluency for their own understanding, can't spot what it gets wrong, and stop building the schema they would have needed.
This compounds both of the failure modes above. Traditional training that produced thin schema produces thinner schema when learners use AI to study. Regulated CE that produced no measurable learning produces no measurable learning faster when AI helps the learner skim through it. And the sponsoring firm has no signal — completion data does not distinguish the licensee who built durable competence from the one who let AI think for them.
The gap, plainly stated. Three failure modes — old-style CBT, hours-driven regulated CE, AI-mediated shortcuts — are all producing the same outcome: licensees and trained workers who have completed something but cannot reliably do the thing the training was supposed to enable. No current platform measures this gap. Sponsoring firms increasingly feel it. Whether regulators eventually formalize a move past completion data is uncertain — that question is one of the open variables this brief names later.
The convergent problem above has a name in cognitive science: the memory paradox. Knowledge moves from declarative recall ("I can name this") to procedural fluency ("I can do this") only through effortful retrieval and practice. The easier we make access to information without doing that effortful work, the less durable the knowledge we actually build. The paradox: the very tools that make information most accessible — search engines, summaries, AI assistants — also most efficiently prevent the brain from doing the work that turns information into transferable skill.1 AI is the most powerful effortless-access tool ever invented, which makes it the sharpest case of the memory paradox the L&D field has faced. Every design choice in the sections that follow is grounded in this science, and that is why the brief uses the phrase as shorthand throughout.
Before getting into what ReadyEngine Phase 2 does, it helps to map the journey it sits inside. A licensee — or any regulated, credentialed, or competency-bearing worker — moves through five stages, not two, and the learning goal at each stage is different.
Build declarative knowledge. Prepare for the gate.
Served by ReadyEngine + ReadyRating today.
Regulator's gate. Discrete moment in time.
ReadyRating pass-guarantee at RR5.
Build schema. Apply firm-specific knowledge. Survive the 0–24 month window where most errors and complaints originate.
Mostly on-the-job, mentor-dependent. Strong Mode 2 fit — firm-led training lives here.
Deepen schema. Maintain competence as products, regulations, and contexts evolve.
CE currently logs hours, measures nothing.
Specialization, novel cases, transfer to complex situations.
No platform currently serves.
ReadyEngine today serves Stage 1, optimized for the moment of Stage 2 — and serves it well. Phase 2 is the platform for Stages 3, 4, and 5 — the stages the regulator's stated goal is built around, but that no current measurement infrastructure addresses.
The regulator's stated goal — published in the competency profiles maintained by CISRO, CIRO, RECA, and their peers2 — is that the licensee moves through all five stages competently. The exam is their gating mechanism for Stage 2 and, by design, a partial proxy for everything in the profile that follows. The published competency profiles themselves demonstrate this: they include ethics, judgment, client-relationship, and contextual reasoning competencies that no multiple-choice exam can reliably assess. Firms and senior practitioners feel this gap in everyday practice. What is missing is a platform that actually serves the full goal, not just the gate.
Stage 3 is where firm-specific knowledge starts to matter more than regulator-published curriculum. Products, processes, client patterns, internal standards, the firm's own playbook — none of this is in any exam blueprint, and none of it is in any regulator's CE catalog. It lives in the senior practitioners' heads, in the firm's incident archives, and in the case studies that come out of the firm's own work.
Sponsoring firms have always known they need to do something here. What they have not had is a platform that combines their firm-specific content with the pedagogical structure that actually builds durable schema. They run onboarding programs, mentor pairings, lunch-and-learns, internal case reviews — all of which produce variable outcomes and zero longitudinal data. Stage 3 is also the developmental window where AI in the workplace is currently doing the most damage to schema formation, because junior practitioners reach for AI before their own thinking has consolidated.
This makes Stage 3 the natural inflection point for Mode 2. The firm brings the content (their case studies, their incident archives, their internal standards); the platform brings the pedagogy and the cognition telemetry. Stage 3 is where the platform's value to the firm becomes most concrete — and where the data we accumulate is the data the firm cares most about, because it's the period that produces or prevents the next complaint, the next compliance incident, the next quiet attrition.
The strategic frame. Phase 1 (ReadyRating) is the gate. Phase 2 (ReadySignal) is the goal. Both products live inside ReadyEngine. Together they cover the regulator's stated objective — competence across the full career arc — not just the moment the regulator currently measures. And Stage 3, where firm-led training has always lived, is the place where Mode 2 most obviously belongs.
ReadyEngine Phase 2 — for now, ReadySignal — is the platform layer that addresses Stages 3 through 5 of the competence journey. It is built on a single premise:
Real professional competence is built only through effortful, schema-deepening, transfer-tested practice. The platform's job is to deliver that practice and to measure the cognition it produces.
That premise inverts almost every dominant pattern in the L&D market. Today's platforms push content at learners and measure consumption. ReadySignal pushes problems at learners first — the learner does the cognitive work before any teaching arrives — and measures cognition (schema density, retrieval strength, transfer capacity, calibration, productive-struggle engagement, practice independence). Content becomes a response to what the learner's attempt revealed, not a precursor to a test of whether they absorbed it.
Every learning interaction starts with the learner attempting something hard before any scaffolding shows up. Difficulty is deliberately held at the edge where roughly one in six attempts fails — the productive-struggle zone where neural growth happens.4 Concepts return weeks and months later in novel surface contexts, not as repeated drills. Mastery is demonstrated through transfer to situations the learner has not seen before. The data layer surfaces something nobody else in the L&D market currently surfaces: a real, longitudinal signal of professional cognition development.
The clearest way to understand what ReadySignal is for is to look at the difference between the two ways a learner can finish a piece of training.
Drawing on Anderson's declarative/procedural memory framework and Bjork & Bjork's desirable difficulties.14
This is where the commercial frame opens up. ReadySignal is not one product for one buyer — it is a platform with two distinct deployment modes. The same architecture, the same pedagogy, the same cognition telemetry layer, but two different commercial postures depending on who supplies the content.
Mode 2 is the bigger market and the more interesting strategic move, so it deserves its own explanation.
A licensing customer — say, an industry association that runs safety certification training for 50,000 workers a year, or a large engineering firm with its own internal training program — comes to us with their content already written. What we provide is everything else: the scenario delivery engine, the productive-struggle calibration, the cognition telemetry layer, the data dashboards, the audit trail. They keep their content, their brand, their learner relationships; we provide the infrastructure that makes their training measurable.
What they need from us to make this work:
A content authoring interface. Not raw HTML or a third-party LMS workaround — a purpose-built tool that lets their instructional designers create branching scenarios, transfer assessments, retrieval items, and reflection prompts using the pedagogical patterns the platform expects. The platform produces good signal only when the content is shaped the right way; the authoring tool teaches the designer how to shape it correctly while they're building.
An ontology mapping layer. Their content has to be tagged against a competency profile — either one they bring or one we help them build. This is what lets the telemetry roll up to meaningful signals at the competency level rather than just the item level.
Their existing identity infrastructure. SSO, learner records, course-catalog integration — the platform sits inside their existing learning ecosystem, not as a parallel system the learner has to log into separately.
Dashboards in their brand. The L&D leader and their executive sponsors see the cognition telemetry through the customer's brand surface, not ours. The platform is plumbing; their training is the product.
The authoring tool supports multiple content-creation pathways, but the most powerful — and the one we would lead with for new Mode 2 customers — is case-study upload. A firm's L&D team, subject-matter expert, or senior practitioner uploads a real case from their own practice: an actual client situation, an incident report, a project that went well or badly, a regulatory inquiry that taught them something, an onboarding scenario new hires keep stumbling on. The platform asks structured questions about the case — what was the decision point, what were the realistic alternatives, what's the underlying competency, what would a different decision have produced — and transforms the case into a branching scenario with productive-struggle calibration, retrieval scheduling, and Socratic prompts baked in.
This is the lowest-friction way to onboard a firm onto Mode 2. They do not have to write content from scratch or translate an existing course into a new format. They bring the cases they already have — the institutional knowledge sitting in their senior people's heads, their incident archives, their anonymized client files — and the platform handles the pedagogical structure around them. Every case they upload becomes both training material for their workforce and a contribution to their firm-specific competency profile and longitudinal cognition data.
The pedagogical insight underneath this is that real cases produce better schema-building than synthetic ones, because they carry the messy ambiguity, contextual nuance, and consequence-relevance that learners' brains actually engage with. A firm's own cases are also aligned to that firm's competency profile by construction — they happened in that firm, to that firm's people, with that firm's products and clients in play. This is why case-study upload is not just a convenience feature: it produces better learning than course content authored from scratch by an outside vendor, because it teaches what actually matters in that specific firm.
Why this feature matters strategically. Case-study upload turns a firm's accumulated experience — the most expensive learning their senior people have done — into reusable training infrastructure for the next generation. It is the most direct line between "we know things we cannot pass on" and "we have a platform that passes them on, measurably." For most enterprise customers, this will be the first feature we demo and the first one they understand.
What we get from them: data. Every cohort that runs through their content on our platform produces telemetry that strengthens the underlying models, validates the cognition signals, and accumulates into the longitudinal data asset that makes the platform more defensible over time. Mode 2 is where the data moat compounds fastest, because it scales horizontally across many content domains, not just regulated CE.
The strategic point. Mode 1 is the natural extension of ReadyEngine; Mode 2 is what makes this category-defining. The first is selling a product into a market that already exists. The second is creating a market — for cognition diagnostics applied to whatever content you happen to be teaching. Both can run in parallel from the start, but Mode 2 is where the commercial ceiling is meaningfully higher.
One of the most important questions inside WKT about what Phase 2 is for is: shouldn't every job have a competency profile we are writing to, rather than an exam blueprint? Yes — and the distinction is exactly the strategic spine of Phase 2.
In regulated professional licensing, three documents exist (or in some cases, are deliberately collapsed):
ReadyEngine Phase 1 (ReadyRating) writes to the exam blueprint — correctly, because its job is exam preparation. ReadyEngine Phase 2 (ReadySignal) writes to the full competency profile. That includes the elements the exam couldn't test — the judgment, the ethical reasoning, the client-relationship work, the response-under-pressure that defines actual practice.
This distinction matters for three reasons.
It is what the regulator actually wants. Every regulator we've looked at has a stated goal that licensees are competent across the full profile, not just at the exam subset. The exam is their measurement tool, not their objective. A platform that demonstrably writes to the full competency profile delivers what they say they want.
It answers the question for non-licensed contexts. Outside regulated professions, there is no exam blueprint. There is only the competency profile — what the role requires, whether published by the employer, by an industry association, by a safety body, or built collaboratively. ReadySignal works exactly the same way in non-licensed contexts: write to the competency profile, deliver the practice, measure the cognition. The platform is regulator-agnostic; the content discipline travels.
It is what unlocks the Mode 2 licensable-platform business. A licensable diagnostic platform that anchors to the competency profile can be deployed across any training domain that has — or can produce — a competency profile. Safety training has one. Compliance training has one. Skills training has one. Each becomes a potential market for the platform.
The variation in how regulators currently structure these documents is worth knowing about, because it affects how Phase 2 lands in each market.
| Regulator | Document structure | Where Phase 2 fits |
|---|---|---|
| CISRO (LLQP) | Competency profile, exam blueprint, and curriculum kept distinct — page 4 of the profile names the distinction explicitly. | Strongest Phase 2 case. The profile is rich and explicitly broader than the exam; we write to the full profile as the regulator already intends. |
| CIRO | Three documents maintained per role, kept separate. Pre-licensure is exam-prep; firms expected to provide additional competence development. | Phase 2 fills the firm's competence-development obligation with measurable evidence. The platform completes what CIRO's architecture already assumes. |
| RECA | Documents collapsed — exam blueprint, curriculum, and competency profile effectively the same artifact. | Phase 2 has to make the case that the role's true competency profile is broader than the regulator's collapsed document. Real opportunity to define what a richer profile looks like. |
| CIRE | No published profile; competence largely defined operationally by the employing institutions. | Phase 2 partners with the bank or credit union to define the profile, then writes to it. Mode 2 deployment. |
In every case, Phase 2's job is to write to the competency profile — published, partial, or built. That is the strategic anchor, and the answer to the question your leadership has been asking.
Six concrete moves operationalize the memory paradox into a working platform. Each one is something we can build with capabilities ReadyEngine already has the shape of — pending the Phase 0 capability audit's confirmation of what's actually reusable.
Every learning object opens with a prediction or decision task. The learner has to commit to an answer before any teaching content appears. Then the platform responds based on what their attempt revealed. This is "predict first, then compare" as a content architecture, not a feature.
Adaptive difficulty calibrated to productive struggle, not learner comfort. The platform deliberately holds the learner at the edge where roughly one in six attempts fails — the difficulty zone where neural growth and schema-formation happen. Most adaptive engines optimize away from this; we optimize toward it.
Concepts surface again weeks and months later — but in novel surface contexts, not as repeated drills. The retrieval schedule connects ideas across the curriculum so the learner builds a connected web rather than isolated facts. This is what distinguishes Phase 2 retrieval from Phase 1's exam-day retrieval.
The real test is not "can you recognize this question type" but "can you apply this concept to a situation you have never seen before." Periodic transfer assessments — purpose-built branching scenarios — measure whether learning is generalizing. This is the assessment most closely tied to on-the-job performance, and the most expensive content to produce well.
The most powerful retrieval method is making the learner explain a concept — and defend their reasoning — to an AI configured to probe rather than answer. AI as Socratic partner: asks for justification, plays devil's advocate, refuses to provide the answer. The learner does all the cognitive work; the AI is structured friction.
High-fidelity branching cases — a difficult client conversation, a workplace incident, an ethical dilemma, a compliance edge case — with simulated consequences, the option to replay, and progressively harder variants. The scenario library is the production heart of the platform and is significant new work to build to the volume and quality Phase 2 requires.
The design pattern in one sentence. Don't replace thinking. Provoke it.3 Every product decision in ReadySignal can be tested against that line. If the feature replaces thinking, it does not belong in the platform. If it provokes thinking, it does.
This is the most counterintuitive design choice in ReadySignal and the one that creates the most product distinctiveness. Most adaptive engines treat difficulty as a UX problem to be smoothed away. We treat it as a design parameter calibrated to a specific target — the zone where the learner is failing often enough to grow, but not so often that they disengage.
Bjork & Bjork's "desirable difficulties" zone — the design target ReadySignal is built around.4
Operationally, this means the platform's adaptive engine carries a different optimization function. Not "keep the learner moving forward at maximum throughput," not "maximize completion rates," but "keep the learner at the edge where measurable cognitive growth is observable in the telemetry." That is a fundamentally different engineering target, and one that no current competitor is even trying to hit.
This is the most natural objection to anything in this space right now: a learner can already ask ChatGPT to quiz them on suitability assessment, generate a practice scenario, or critique their reasoning. Why would a firm pay for a dedicated platform when AI can do most of this?
The objection is worth answering directly, because the answer reveals what the product is actually selling.
It's not the content delivery — it's the system architecture and the data it produces. AI can answer learner questions and generate practice items on demand. What it cannot do, sitting on its own:
Calibrate difficulty deliberately. A learner using ChatGPT chooses what they ask about and how hard to make it — almost always defaulting to comfort. ReadySignal's 85% engine is an external system that holds the learner at the edge of productive struggle, where they would not voluntarily choose to be. AI alone defaults to the wrong difficulty zone for learning.
Track schema growth across time. AI can quiz a learner today. It can quiz them again next week if they remember to come back. It cannot — and does not — produce a longitudinal cognition signal showing how their reasoning, retrieval, and transfer capacity have developed over six or twelve months across multiple competencies, surfaced as data the firm can act on. That signal does not exist unless someone is collecting it deliberately.
Schedule retrieval in novel surface contexts. Spaced retrieval in genuinely novel scenario contexts requires curated content and a scheduling layer. AI generating a fresh prompt each session does not reliably produce the right scenarios in the right sequence to build durable schema. It can mimic the format; it does not produce the pedagogical structure.
Guarantee quality on regulated content. AI hallucinates. In regulated practice — financial services, real estate, healthcare adjuncts, workplace safety — a single hallucinated regulation or invented citation can produce real compliance and liability exposure. WKT-curated content (Mode 1) is reviewed by subject-matter experts and signed off; Mode 2 customers' content is reviewed by their own SMEs in the authoring tool. AI alone has no SME oversight built in.
Produce regulator-grade audit trails. A firm cannot satisfy a CISRO or CIRO audit with screenshots of a licensee's ChatGPT history. They need attested completion records, validated learning evidence, and a chain of custody on the assessment data. ReadySignal produces this; an AI conversation does not.
Hold institutional accountability. The firm has obligations to its regulator, its insurers, and its board. Those obligations require a system of record, not a learner's individual habit of using a chatbot. Even a highly motivated licensee using AI well is generating no institutional evidence of capability — that evidence has to be produced by a platform the firm controls.
The right way to think about it. AI is a powerful general-purpose tool. ReadySignal is purpose-built infrastructure for a specific job — building and measuring durable professional cognition at scale, with the data discipline, content quality, and audit trail that regulated practice requires. AI shows up inside ReadySignal as the Socratic partner that probes the learner's thinking; it cannot substitute for the platform around it. The question isn't "AI or ReadySignal." It's "what does AI sit inside?"
Both products live inside ReadyEngine. ReadyRating (Phase 1) is the existing exam-readiness product; ReadySignal (Phase 2) is what's proposed here. The differentiation is not about platform infrastructure — it's about what the platform is optimizing for, what it anchors to, and what stage of the journey it serves.
| Dimension | ReadyRating (Phase 1) | ReadySignal (Phase 2) |
|---|---|---|
| Outcome it optimizes for | Pass the licensing exam | Build durable, transferable professional cognition that holds up on the job |
| The "test" it is built around | A regulator's exam blueprint at a fixed point in time | A novel professional situation at an unknown future point |
| What it anchors to | The exam blueprint | The full competency profile |
| Stages of the journey it serves | Stages 1–2 (pre-licensure and the exam event) | Stages 3–5 (early practice, ongoing practice, advanced practice) |
| Unit of content | Concept item, practice exam question, retention drill | Scenario, decision point, work-sample task, teach-it-back exchange |
| Assessment philosophy | Match the exam format (high external validity for that exam) | Reject the exam format (because the job is not multiple-choice) |
| How AI is used | Adaptive routing, content generation, mastery diagnostics | Socratic partner explicitly configured not to give answers; AI as friction, not assistance |
| What is measured | ReadyRating composite (5 signals tied to exam readiness) | Cognition telemetry (6 signals tied to professional development) |
| Cadence | Concentrated study period ending at exam day | Ambient and continuous — minutes per week over months and years |
| Deployment | WKT-published, regulator-accredited (Mode 1) | WKT-published in regulated markets (Mode 1) + licensable platform in non-licensed markets (Mode 2) |
The first draft of this brief overclaimed about how much of ReadyEngine's existing infrastructure is directly reusable in Phase 2. The honest version is more uncertain — and the Phase 0 capability audit is what tells us where the truth actually lies.
What likely transfers: the concept ontology itself (what concepts exist and how they relate — the competency profile contains the exam blueprint, so Phase 2 extends rather than replaces); the declarative knowledge content (definitions, rules, regulatory frameworks — the memory-paradox literature says this is the necessary precursor to procedural fluency, so what Phase 1 builds is exactly the foundation Phase 2 needs to layer on); some richer case studies where they exist (LLQP has these; RECA less so).
What likely doesn't: the assessment format (multiple-choice items optimized for exam mimicry are not the right shape for transfer assessment); the exam-day calibration (Phase 2 is not optimizing for a single high-stakes event); the completion / throughput metric (Phase 1 measures progress to readiness for a fixed event; Phase 2 measures ongoing capability development).
What we don't yet know — and what the audit will answer: how the ontology development pipeline scales when we go from exam-blueprint-anchored to competency-profile-anchored content; whether the existing content authoring pipeline can produce scenario content at the volume and quality Phase 2 needs (this is a new content discipline at meaningful scale, and the production economics are an open question); what the integration cost is between the two phases — handing off learner state, ontology connections, cognition telemetry continuity.
The strategic implication: Phase 2 inherits a meaningful foundation from Phase 1, but it is not a wrapper on Phase 1. It is a different product built with shared infrastructure, the precise shape of which the audit clarifies before any timeline is committed.
The clearest way to see where ReadySignal sits is to look at the ReadyEngine platform as it exists today and as it could exist with Phase 2 added.
Same platform. Same data discipline. Different developmental moment in the learner's life.
The two products share a customer arc. A candidate is sponsored through pre-licensure on ReadyEngine, takes their exam, receives their license. The sponsoring firm then continues the relationship into ReadySignal for the same person's post-licensure development. The data the firm sees evolves: from are they on track to pass? to are they developing the cognition the role actually requires? Same platform family. Same data discipline. Same brand. Different developmental question, answered with the same architectural pattern.
The clearest way to see ReadySignal is to walk through what an engagement looks like for different kinds of learners. The platform is ambient and continuous in some contexts and concentrated in others — the engagement model adjusts to the role and the regulatory shape of the training. Three personas illustrate the range.
Maya is eighteen months into her career as a licensed advisor, sponsored by her firm into ReadySignal as part of their advisor development program.
Total engagement: about 10–15 minutes a week. The platform fits into Maya's working life, not pulling her out of it.
Devon is a safety officer at a mid-sized industrial firm. His Working at Heights certification expires every three years. Historically that means an 8-hour refresher course, mostly review of material he already knows, ending in a 30-question multiple-choice test. In and out. The firm logs the certification; nothing else changes.
On ReadySignal, his certification cycle looks different. The renewal still concentrates around a short engagement window — there is no expectation of months of practice for ticket training. But the structure of those hours changes, and a small ambient layer surrounds them.
The selling story for ticket training is not "12 months of ambient engagement." It is "the certification you already have to deliver, made measurable — plus a low-touch retention layer afterwards that prevents the typical decay." Total post-event time investment from Devon across the renewal cycle is under an hour. The data signal it produces is something the firm has never had before: evidence that the certification training actually built durable competence, not just satisfied an hours requirement.
Anya runs L&D for an engineering services firm with 12,000 workers across multiple jurisdictions. The firm has its own competency profiles for every role, built over years with their technical leadership. They have proprietary content — case studies drawn from their own projects, internal standards, hard-won institutional knowledge. They are not going to hand this content over to a vendor.
What they license from us: the platform. Their instructional designers author their existing content into the ReadySignal authoring tool, mapped against their competency profiles. The platform delivers it using the six design moves, produces the cognition telemetry, and surfaces dashboards in Anya's brand. The firm keeps full ownership of its content and its data.
What Anya now has that she didn't before: a real signal on whether her workforce is actually developing the competencies the firm has defined. Not "completed the training" — but "schema is densifying across our critical safety competencies; calibration is improving in the under-3-years cohort; transfer is weak in two specific competency areas where we need to redesign content." The platform tells her what's working and what isn't, with data she can take to her CEO.
This is the Mode 2 buyer in concrete form. The platform is plumbing for her training program. The diagnostic is the product. Across 12,000 workers, engagement is ambient — a few minutes a week, woven into existing role responsibilities — and the data accumulates into something no L&D function has ever had access to.
The cognition telemetry rolls up differently for different roles. Here is what each persona's L&D leader actually sees.
Across all three, the underlying shift is the same: from a completion record to a capability record. The L&D function gets, for the first time, a defensible answer to the question the board has started to ask — do we actually know what our people are capable of?
This is the part that turns ReadySignal from a "better CE platform" into a defensible commercial product. The data the platform produces is something the market has not seen before, and it is what makes sponsoring firms and enterprise customers willing to pay for it.
Six signals make up the cognition telemetry layer. Each one is something the platform can actually observe, each one is something firms genuinely want, and each one is something no current competitor measures.
Concepts connected to ≥3 other concepts. Up from 52% twelve months ago. Cohort average: 68%.
Concepts retrieved correctly at 3-month delay in novel surface context. Trending up.
Performance on previously-unseen scenarios using familiar concepts. Cohort avg: 61%.
Confidence aligned with capability. Knows when she knows; flags when she doesn't.
Stays at the edge instead of avoiding hard scenarios. Strong motivational signal.
Needs 27% less scaffolding than 12 months ago to solve cases at equivalent difficulty.
Schema density. How richly the learner's concepts are connected to each other. Visualizable as a knowledge graph that gets denser over time. The closest direct measure we have of "they can think with this material" as opposed to "they have memorized this material." Pointer learning produces thin graphs; schema learning produces dense ones.
Retrieval strength. Durability over delay. Can the learner pull this concept back two weeks later? Three months? A year? In a novel surface context, not as a repeated drill? This is the gold-standard cognitive-science measure of durable learning and what predicts on-the-job availability of knowledge under pressure.5
Transfer capacity. Performance on novel scenarios that use familiar concepts. The hardest signal to measure but the closest to the thing firms actually want — does the learning translate to situations the person has not seen before? The only signal that predicts job performance directly.6
Calibration. Do they know when they know? Over-confidence is a measurable risk signal — the advisor who is sure they understand replacement-insurance disclosure but is actually wrong is exactly the advisor who will generate the next complaint. Well-calibrated uncertainty is a competence signal.7
Productive struggle engagement. Is the learner choosing to stay at the edge, or are they avoiding difficulty by sticking to easy scenarios? A motivation-and-grit signal nothing else in the L&D market currently captures, and highly predictive of long-run development trajectory.
Practice independence. Over time, does the learner need less scaffolding to solve cases at equivalent difficulty? The operational measure of expertise growth — the trajectory from novice-needs-support to expert-works-unaided.8
Why this is commercially defensible. Every signal above is computable from telemetry the platform can actually observe. None require self-report. None require subjective scoring as the only input (rubric grading on transfer scenarios is one input, but the rubric is well-defined and inter-rater reliability is testable). And — critically — every signal accumulates value over time. A customer that has been using ReadySignal for three years has data nobody else can replicate, because the data only exists if you have been observing the workforce for three years.
Five distinct buyer profiles, mapped across Mode 1 and Mode 2 deployment.
Who: Insurance carriers and MGAs sponsoring LLQP-licensed advisors. Investment dealers sponsoring CIRO-licensed reps. Real-estate brokerages sponsoring RECA-licensed agents. Banks and credit unions sponsoring CIRE-licensed CSRs. Any regulated employer sponsoring a licensed workforce in Canadian financial services or adjacent industries.
What they need: Defensible evidence that their licensed workforce is developing the professional cognition the role actually requires. Audit-ready CE compliance with substance underneath it. Pre-complaint risk signals. Onboarding intelligence for new hires. A defensible answer to the AI-in-the-workforce question their board is starting to ask.
What they currently buy and why it doesn't solve the problem: Hours-based CE catalogs (tick-the-box compliance, no developmental signal); LMS platforms (track completion, not capability); occasional in-house assessments (expensive, episodic, not longitudinal).
What they would pay for ReadySignal: Per-seat annual licensing tied to their CE cycle, structured as a workforce-development line item rather than a training expense.
Who: Enterprise L&D functions in banks, insurance carriers, healthcare networks, energy companies, transportation firms, government departments. The internal team that owns workforce development and reports to a Chief People Officer or General Counsel.
What they need: A defensible answer to the board question "do we actually know what our people are capable of?" Evidence of L&D ROI in a form that survives scrutiny by Finance and Risk. A way to demonstrate that their training investment is producing capability change, not just activity.
What they currently buy: A patchwork of LMS, content libraries, in-house training, and consulting engagements. None of them produce a longitudinal capability signal at the workforce level.
What they would pay for ReadySignal: Enterprise contracts, with optional content-authoring services bundled in. The data layer is the differentiator; the content is the delivery vehicle.
Who: Workplace safety training providers (Working at Heights, WHMIS, OH&S course providers), certification bodies (Red Seal trades, industry-specific cert programs), and the in-house safety functions in heavy industry that deliver mandatory training internally.
What they need: Evidence that their ticket training actually builds durable capability, not just satisfies a hours requirement. Defensible documentation for insurers — post-incident, the question is often "what did your training actually teach them?" Reduced repeat-incident rates from workers who supposedly received training.
What they currently buy: LMS systems for delivery, content authoring tools for course design, and reporting tools for compliance. Nothing measures actual cognition development inside or after the training event.
What they would pay for ReadySignal: Mode 1 if they want our content packaged with the platform; Mode 2 if they have their own content and want our diagnostic layer. The selling story is durability evidence — what happens to the training months after the certificate is issued. Pricing structured per-credential per worker, or per-program per year.
Who: Large enterprise L&D functions in technology, engineering, manufacturing, energy, healthcare networks, government — organizations with their own competency profiles, their own content, and a workforce-development mandate that has outgrown what their LMS can support.
What they need: A diagnostic layer for their existing training investment. They are not going to switch content vendors or rebuild their L&D function around a new platform — they need infrastructure that lifts what they already have.
What they currently buy: Enterprise LMS, content libraries, consulting engagements. None of these produce a longitudinal capability signal; all of them produce activity logs.
What they would pay for ReadySignal: Platform licensing + per-seat usage, with content authoring tools, ontology mapping, and dashboarding included. This is the highest-ceiling commercial opportunity in the brief because there are tens of thousands of organizations in this category and almost none of them have what we're offering.
Who: CISRO, CIRO, RECA, the provincial real-estate councils, eventually peer bodies in other regulated industries. The organizations that set the CE rules and certify training providers.
What they need (hypothesized): A way to certify CE that produces evidence of development, not just hours. The current CE model is widely understood inside the L&D industry to be a weak proxy for actual capability. Whether regulators will formally move toward outcome-based certification — and on what timeline — is one of the questions the Planning & Validation phase has to answer through direct regulator engagement. The premise of this buyer profile is that they will, eventually; that premise needs to be tested, not assumed.
What we would offer: Certification of ReadySignal as an "outcome-based CE" provider, with the cognition telemetry serving as the evidentiary basis. A long-horizon play (regulator timelines are slow), but it is what makes ReadySignal category-defining rather than category-occupying.
The strategic significance: If even one major Canadian regulator certifies outcome-based CE as a recognized model, the market shifts. Every CE provider then has to compete on capability evidence, and we are the only one with the data layer built.
Five things make ReadySignal defensible, and they layer on top of each other.
1. Pedagogical IP — memory-paradox-native design. The product is built on a set of design principles most edtech competitors will not embrace because they cut against the dominant UX instinct of the L&D market (frictionless content delivery, completion optimization, comfort-first adaptive engines). Anyone who copies our content without copying our design will get an inferior product. The principles are simple enough to articulate and counterintuitive enough that most won't implement them.
2. Measurement IP — the cognition telemetry layer. The six signals, and the algorithms that compute them from raw scenario telemetry, are proprietary work. They are also genuinely hard to build well. Anyone trying to replicate this needs both the cognitive-science expertise to define the right signals and the platform engineering to compute them reliably. Few competitors have both.
3. Regulator relationships — the Relo precedent and beyond. ReadyEngine has already built regulator-side relationships that allow it to be accepted as pre-license training in at least one jurisdiction. Those relationships are extensible to CE certification work, but only if you've already done the trust-building work. New entrants will spend years catching up.
4. Longitudinal data — the asset that only accumulates over time. The most defensible thing about ReadySignal is the data it produces. A customer that has been using ReadySignal for three years has cognition-development data on their workforce that no competitor can replicate, because the data literally only exists if you have been observing the workforce for three years. This is a winner-takes-most dynamic in the category.
5. First-mover position in a category that doesn't exist yet. There is no current product called "professional cognition telemetry" or "outcome-based CE platform" or "competency-profile-anchored diagnostic." We define the category. We set the vocabulary. We make the first regulator submissions. Every competitor who follows enters a market we have framed and a conversation we have shaped.
The mental model. Think of ReadySignal like a cousin of a credit bureau, but for professional capability. The firms that win in credit-scoring are not the ones with the best algorithm — they are the ones with the deepest historical data. The same dynamic applies here. The longer we operate, the more defensible we become. The earliest customers we win become the data foundation everything else is built on.
This is a meaningfully complex new-product effort, and the build cannot be reduced to a tidy timeline. The first version of this brief understated that complexity. The honest version, below, names what each phase has to produce and the validation that has to happen before the next phase begins. No durations are committed — each phase ends when its required outputs are real and defensible, not when a calendar deadline arrives.
The principle. The most expensive mistake we could make is moving to MVP before the underlying premise is validated. The pedagogical claims, the cognition signals, the buyer hypothesis, the content authoring economics, the regulatory pathway, the platform reusability — all of these need to be tested before we commit to building. A polished MVP built on an unvalidated foundation is far more expensive than a slower, more deliberate path that gets the foundation right.
On timelines. Deliberately removed from this section. The first draft of this brief committed to durations (90 days for Phase 0, 12 months to MVP) that reflected ambition more than realism. Each phase here is sized by its required outputs, not by a calendar. Once the planning and validation phase produces honest evidence, a defensible timeline can be built around what the work actually takes. Until then, any commitment to a date is premature.
Every claim in the table below is honest about uncertainty. The ALF capability audit (in the Planning & Validation phase) verifies what's actually there before any build cost or timeline gets committed.
| Capability | Hypothesis | Confidence |
|---|---|---|
| Adaptive content engine (ALF) | Reusable as the delivery foundation; needs different optimization target for Phase 2 | High — audit confirms scaling characteristics |
| Multi-signal telemetry infrastructure | Re-purposable for cognition signals; some may be derivable from data already collected | Moderate — audit confirms what's actually instrumented |
| Content authoring pipeline | Scaffolding likely transfers; scenario-specific extensions and authoring tool for Mode 2 are new build | Moderate — audit clarifies capacity and gaps |
| Scenario authoring (branching narrative) | New build at the volume and quality Phase 2 requires; WKT has scenario design experience but not at this scale or shape; integration with ALF data layer also new | Low — significant new investment needed |
| Ontology / concept model | Exists for exam blueprints; needs extension to competency profiles | Moderate — audit clarifies extensibility |
| Regulator relationships | LLQP relationship extensible; other regulators take longer | High |
| Brand position | Readiness Intelligence extends naturally; Phase 2 product name still TBD | High |
| Socratic AI interaction layer | New build; prompt engineering + UX, not novel ML research | High |
| Cognition telemetry algorithms | New build; cognitive-science design + statistics | Moderate — some signals more tractable than others |
| Schema-mapping data model | Requires new cross-concept tagging during authoring; ongoing investment | Moderate |
| Mode 2 content authoring tool | New build; substantial product engineering for licensee-facing UX | Tractable but non-trivial |
| Firm-facing dashboard UX | New build; standard product engineering | High |
| Outcome-based CE certification pipeline | New regulatory work per jurisdiction; tractable but slow | Moderate — depends on regulator |
The pattern: most of the platform is buildable on a tractable timeline by a team WKT already has the shape of. What was overclaimed in the first draft was the level of direct reuse from ReadyEngine — that is still likely substantial, but the audit is what tells us, not assumption.
The platform name (ReadyEngine) and the Phase 1 product name (ReadyRating) are fixed. What's open is the Phase 2 product name. Two candidates worth comparing.
Strengths: Carries the cognition-telemetry framing directly in the name — the platform produces a signal that didn't exist before. Works equally well across Mode 1 (regulated CE) and Mode 2 (licensable platform) — both modes produce the signal. Extends the Ready- brand line cleanly. The most distinctive thing about the product is the data it generates, and the name says exactly that.
Risks: Less actionable than ReadyPractice — describes what the product produces rather than what the learner does. Some buyers will need brand explanation. Mitigation: brand discipline in messaging keeps "Signal" as the data layer; "Practice" can be a sub-brand or feature name for the engagement layer.
Strengths: Clean extension of the Readiness Intelligence brand. "Practice" carries the dual meaning of professional practice (what they do at work) and learning practice (how the platform builds capability). Buyer-friendly; easier to explain to non-L&D audiences. Strong fit if we believe the engagement layer is the most distinctive thing.
Risks: Could be confused with "practice exam" in the candidate-prep context. Less directly evocative of the data/measurement layer that is the platform's deeper commercial story. Stronger as a feature name within ReadySignal than as the parent product name.
Recommendation: ReadySignal. Given how central the cognition telemetry layer is to what makes this product defensible and category-defining, the name should carry that frame. ReadyPractice describes what the learner does; ReadySignal describes what the platform produces. The data is the differentiator, and the name should say so. ReadyPractice remains useful as a sub-brand for the engagement layer within ReadySignal — but as the primary product name, ReadySignal is the stronger choice.
A concept brief is worth more if it names what it does not yet know. Eight open questions, in rough order of how much they matter for the go/no-go decision.
The single biggest unknown. The brief has been deliberately careful about claims of direct reuse, because the truth of those claims has not been verified. The Phase 0 audit — which inventories ReadyEngine's actual instrumentation, authoring capacity, data model, and infrastructure — is what determines the build cost and timeline for Phase 2. If the audit reveals significant rework is needed, Phase 2 is more expensive and slower. If it confirms strong reusability, Phase 2 is faster than the roadmap suggests. The audit is load-bearing for everything downstream.
The hypothesis is that the cognition telemetry layer is the commercial differentiator and firms will pay materially more for it than they pay for current CE. The alternative is that firms will buy on compliance grounds (cheaper, faster, easier audit) and treat the cognition data as a nice-to-have. Phase 0 customer conversations resolve this — and the answer changes the business model meaningfully.
The Mode 2 commercial opportunity is much larger than Mode 1, but it is a market we're hypothesizing rather than a market we know. Enterprise L&D buyers will respond to the diagnostic value proposition in principle, but they are hard to sell to and slow to procure. Phase 0 buyer discovery has to include 4–6 Mode 2 candidate customers to validate that the buying signal is real and the price-point is achievable.
The six signals are theoretically grounded and operationally well-defined, but they have to actually predict on-the-job performance for the moat to hold. The Phase 2 validity research — correlating cognition signals with workplace outcomes — is the load-bearing piece of evidence. If the correlations are weak, the data layer is less defensible. Resolved by Phase 2 published research.
Ticket training has structural pull toward shortness, simplicity, and predictability. The selling story — "your certification cycle, made measurable, with a low-touch retention layer" — has to overcome the in-and-out mentality hardwired into the buyer's expectations. The risk is that even when the value proposition is clear, behavior change is slow. Mitigation: lead with cohort-effectiveness dashboards (the buyer wants those even without changing what learners experience) and let learner-facing changes follow.
The regulatory timeline is the slowest variable in the model. CISRO, CIRO, RECA each move on their own clocks. If certification takes 5+ years across the board, ReadySignal can still operate as a development product that also happens to satisfy current CE rules — but the category-definition story is slower to mature. Resolved gradually through Phase 2 and Phase 3 regulator engagement.
The design depends on AI that probes rather than answers. There is real risk that learners dislike the friction (especially compared to ChatGPT-style answer-givers they're used to) and that engagement metrics suffer in early testing. Mitigation: careful UX design — clear scaffolding around the Socratic interaction, transparent labelling of the AI's role, and a fall-back "explainer" mode for moments when the learner is genuinely stuck.
Cognition telemetry is sensitive — essentially a workforce capability profile, with employment and labour-relations implications. The design choice we recommend: the learner owns the data, with consented sharing to the sponsoring firm in aggregate (cohort patterns) by default and individual sharing only with affirmative consent. Defensible under Canadian privacy law and most provincial labour codes, but the design has to be deliberate and the legal advice has to be solid.
None of these risks is a deal-breaker. Each is a real question to answer, and each has a credible answer-path. The combined risk profile is consistent with a serious new-product development effort in an adjacent category — not a moonshot, not a sure thing, but a tractable bet with a clear methodology for resolving uncertainty stage by stage.
Three failure modes are converging in the L&D market right now: traditional content-driven training that produces shallow, decay-prone learning; regulated CE and ticket training that measure hours and not capability; and AI in the workforce that compounds both problems by substituting fluency for the schema licensees never built. The current platform market — LMS, content libraries, AI tutors, CE catalogs — is not designed to address any of these.
ReadyEngine Phase 2 (working name: ReadySignal) is the platform layer designed to address them. It uses memory-paradox-native pedagogy — retrieval-first content, productive-struggle calibration, schema-building retrieval across novel contexts, transfer assessment, Socratic AI partners, branching safe-practice scenarios — to deliver professional development that actually changes professional capability. It measures that capability development through six cognition signals no other platform currently produces. It anchors to the full competency profile (what the role actually requires) rather than the exam blueprint (what the regulator's current gate happens to test) — which is what the regulator's stated goal already wants. And it deploys in two distinct modes: as a WKT-published product for regulated CE markets we know, and as a licensable diagnostic platform that other organizations bring their own content into.
The two products inside ReadyEngine share a customer arc, a platform foundation, and a brand. Phase 1 (ReadyRating) gets the licensee through the regulator's gate. Phase 2 (ReadySignal) builds and measures the competence the gate was a proxy for. Together they cover Stages 1 through 5 of the competence journey — what the regulator says the licensee needs to be, not just what they currently get measured on.
The build is sequential, each phase ending when its outputs are real and defensible rather than at a fixed date. The first and longest phase is planning and validation — expert advisory panels (cognitive scientists, psychometricians, L&D experts), construct-validity work on the six cognition signals, an honest ALF capability audit, a content authoring feasibility study, deep buyer discovery, business model design, regulatory pathway analysis, and legal framework design. Only once that foundation is established does the platform move into prototype, then pilot, then MVP, then scale. The work is more complex than a tidy roadmap suggests, and any timeline committed before the validation work is done is premature. The moat is five layers deep: pedagogical IP, measurement IP, regulator relationships, longitudinal data asset, and first-mover category position. The buyers are five distinct profiles across Mode 1 and Mode 2 — sponsoring firms, internal L&D, safety and certification providers, enterprise platform licensees, and regulators on the longer arc.
The strategic ambition is to define a new category — outcome-based, cognition-measured, competency-anchored professional development infrastructure — before any competitor names it. If we move now, we set the vocabulary, frame the conversation, and accumulate the longitudinal data that becomes the asset other entrants cannot replicate.
— Emma Plumb
Prepared for WKT leadership · Emma Plumb · May 2026
Working name "ReadySignal" used throughout — ReadyPractice is the alternative; naming decision in §Naming Phase 2.
This brief stands alongside the ReadyEngine Strategic Brief (May 2026) as a parallel concept exploration.