Real-time competency assessment for electrical installation training. AI-generated evidence packs, auto-mapped to City & Guilds. Zero chasing signatures.
They became trainers because they're excellent at the trade. This is what their admin actually looks like.
The only part that actually matters.
Often done hours later. The details matter at audit. The details get lost.
80-page PDFs. Updated constantly. Always the wrong tab open.
Different format from the awarding body. Different from internal records. Always.
They're in a different building. They're on a visit. They'll get to it.
"Just in case" an audit comes. It always comes.
The trainee starts the task. Cognara's assessment engine watches for technique, sequence, safety protocol, and completion quality โ mapped to your specific framework in real time. No setup. No coaching the AI. It knows your course. Detection overlays track hands, tools, and safety equipment as each criterion is assessed silently in the background.
You show them a 90-second clip of a trainee doing an electrical task. The Cognara interface populates alongside it โ criteria ticking, evidence clips auto-tagging, timestamps appearing. Then you ask: "How long does that normally take you to write up?"
They say: "About an hour. Maybe more."
You say: "This was done before the trainee put his tools down."
Vision AI can now reliably interpret procedural, physical tasks in real workshop environments. Not just faces in clean studio conditions โ but hands doing specific things with specific tools in noisy, variable, industrial settings.
And LLMs can now map unstructured observed behaviour onto formal competency frameworks without a human translator in the middle. Accurately. Consistently. Compliantly.
Neither was reliable enough to ship 18 months ago. Both are now. That's the window. Training providers who move first build an evidence archive their competitors can't replicate.
Face recognition, basic object detection. Workshop environments: unreliable. Procedural task assessment: impossible. Too many false positives in real-world conditions.
Framework mapping theoretically possible. In practice: error rates too high for compliance use. One wrong criterion in a moderation pack and the whole thing unravels.
Vision + LLM combination reaches reliability required for live assessment. Cognara prototype validated with five training providers across three sectors.
Real frameworks. Real audits. Real outcomes. The technology and the compliance context are finally aligned. This is the moment โ and the window won't stay open.
Every competency framework, plain-English mapped, updated automatically when awarding bodies change their standards. Quarterly โ minimum.
You know exactly which assessors are behind on sign-offs, which employers are chasing updates, and which moderation visit is six weeks away. You need evidence you can stand behind โ not evidence assembled at 10pm the night before. Cognara is your operations backbone.
When an employer asks "prove your training delivers outcomes," you need data โ not anecdotes. Cognara gives you a live dashboard of competency attainment across every cohort, every course, every site. Board-ready. Audit-ready. Always.
You became a trainer because you're excellent at the trade โ not because you love paperwork. Cognara handles the documentation so you can spend your time where it matters: on the floor, with the trainees, doing what you're actually good at.
Two assessors were spending over three hours per trainee on documentation. Evidence packs were being submitted late โ regularly. In 2023, a City & Guilds moderation visit raised 14 individual evidence queries. The provider nearly lost their approved centre status.
Cognara went live in February. By March, assessment documentation time had dropped to 18 minutes. By April, the next moderation visit passed with zero queries raised.
"I showed the moderation officer our evidence trail and she asked what system we were using. That has never happened before."
โ Sarah McLean, Training Operations ManagerReal quotes from training professionals using Cognara in live workshops, day to day.
Every training provider that runs Cognara makes our assessment AI more calibrated โ to real workshop conditions, to specific trade movements, to what "correct" actually looks like on the floor versus in a textbook.
A competitor starting today would need to ingest 23 awarding body frameworks from scratch โ thousands of hours of unglamorous mapping work that no VC will fund a second time.
And once a training provider's evidence archive lives in Cognara, migrating it is a compliance liability they will never voluntarily trigger. The switching cost compounds with every cohort.
23 awarding bodies. 14 sectors. Every framework updated quarterly in plain English. A competitor starts at zero. This took us 18 months. It will take them 18 months too.
Every task assessed improves the vision model's understanding of what correct technique looks like in a real, dirty, variable workshop. This data isn't purchasable. It's earned assessment by assessment.
Once a provider's historical evidence archive is in Cognara, it becomes a compliance asset. No one migrates a system that their moderation trail depends on. The moat grows with every enrolment.
Cognara-generated evidence has now passed moderation visits across six awarding bodies. Moderators have seen it. Some have asked about it. That trust takes years. It cannot be purchased.
Five questions. Honest answers. You'll know exactly where you stand.
All plans include a 60-day evidence audit guarantee. If Cognara-generated packs are ever queried in a moderation visit, we fix it โ at no extra cost.
60-day evidence audit guarantee. If Cognara-generated evidence packs are ever queried in a moderation visit during your first 60 days, we'll work with you to resolve it โ at no extra cost. Your data is always yours. Export everything, anytime. Cancel anytime.
Cognara gives them their time back. Every assessment documented before the trainee walks away. Every evidence pack audit-ready before the moderator calls. 6 weeks from signing to first live assessment.