Book a demo
Menu Close

How it works

A course where every exercise ends in a decision you have to defend.

Symerse is hands-on AI training: an instructor runs the room, trainees work in pairs on shared iPads, and every call they make is scored.

Book a demo Read the evidence

Delivered with NIBRT in Dublin since June 2026.

Symerse exercise: the pair's own call, Hold, recorded first, then the AI's call shown beside it with the option to agree or override.
A pair makes its own call first, then sees the AI's.

An instructor, a room and pairs at shared iPads.

Trainees work in pairs, so every decision is argued out loud before it is submitted. The instructor runs the room from a web console and debriefs the cohort after every exercise.

It runs in an ordinary classroom. A GMP suite is not required.

The flagship runs 09:00 to about 16:20 on both days, six exercises a day. The two days, hour by hour

The trainer console room board: every table against every exercise, with each submitted score as it comes in.

Four kinds of exercise, all built on real work.

Symerse exercise with an AI persona, Dr. Dara Quinn, senior process and reliability engineer, answering questions beside the plan for the week.

01Persona consultation

Pairs question AI personas built on real roles, each labelled as AI, who answer from what is on screen and never make the call for them.

Symerse investigation: a fill check table with one reading, 19.62 litres, below the 19.70 litre limit, and a question for the quality lead.

02Investigation

Pairs work through the data to find what is out of specification, then ask the right expert what caused it.

Symerse live-data exercise: a maintenance board ranking jobs by failure risk, with vibration and pressure trends against alarm limits.

03Live-data decision

Pairs read condition and process data, pick what to act on first and test the choice on a model of the plant before committing.

04Capstone

Each pair designs an AI pilot for their own site, scores it on readiness, risk, impact and feasibility, and pitches it to the room in two minutes.

3D model

A 3D bioreactor, with the AI layer switched on.

In one Symerse exercise, trainees work on a 3D model of a 2,000 L bioreactor. Switch on the AI layer and every sensor, AI model and record that feeds the decision appears around the vessel.

It is the exercise in Inside the Golden Batch, a short session run three times so far, around the IDA-NIBRT Summit in October 2026.

Symerse 3D exercise: bioreactor BR-06 with the AI layer on, showing readings for dissolved oxygen, pH, temperature, glucose and viable cells, the feed pump running 120% over its setpoint, and two flags on the batch record.

Every decision is scored, and the trainee sees why.

Symerse debrief: 81 out of 100, Excellent, with sub-scores of 82 analytical depth, 78 evidence quality, 74 safety and quality and 93 decision quality, and written feedback.

01What gets scored

Everything a pair does in the exercise: the questions they ask, what they flag, what they pick and why.

02What the trainee sees

A score out of 100, a band such as Developing or Proficient, the marks behind it and what to try next.

03What the trainer sees

Every table's score on every exercise, live, and who to follow up with after the course.

Scores are assessed by AI against a shared rubric, and the trainer reviews them with the room.

The trainer runs the room from one console.

Prepare, deliver, review and project, in one place. Scroll sideways for each view.

  1. Trainer console, Prepare tab: a checklist before the day, devices, the agenda and the session code.

    Prepare

    Class code, devices, roster and the agenda, checked before the day.

  2. Trainer console, Deliver tab: six tables by five exercises with states and scores, and one table open showing its score, chat and decision.

    Deliver

    Every table on every exercise, with its state and score, live.

  3. Trainer console, Results tab: six tables, 25% submitted, scores from 59 to 72, and a follow-up list.

    Results

    Submissions, score from first exercise to last, and who to follow up with.

  4. Projector view on a dark screen: six of six pairs joined, waiting for decisions, with the exercise list across the top.

    Projector

    A room view for the screen, showing progress without any trainee detail.

  5. Cohort brief

    Before the day: who is in the room by company and role, and what their sites work with, from the trainees' intake answers.

Every trainee leaves with a record of what they can do.

Sample certificate of performance for a fictional trainee: 66 out of 100, Proficient, with a skills profile and an exercise breakdown.

01Certificate

A co-branded certificate with the overall score and a skills profile.

Sample course record for a fictional trainee: each exercise with its score, the written feedback, the decision and the plan.

02Course record

A full PDF record of every exercise: the feedback, the decision, the plan and the score.

Alumni code

Works on any device for 12 months, with the trainee's own conversations and work preserved.

03Alumni access

A personal code to keep practising after the course, on their own device.

Samples from a demo cohort with fictional names. Certificates carry the delivering organisation's name beside ours.

Three ways to run it, and what is live today.

Live

On our iPads, at your site.

We bring the iPads and an instructor, and run a cohort in your training room.

For manufacturing sites

Live

On your own instance, with your trainers.

Institutes run the courses themselves, as NIBRT does in Dublin, and rearrange the exercises in the composer.

For training institutes

In development

Beside a course you already run.

The Companion puts Symerse exercises beside an institute's existing practical courses.

Questions training leads ask us.

Do trainees need to know anything about AI before the course?

No. Most trainees on the flagship course have never had a conversation with an AI model before the first exercise, so that exercise is an open conversation with no mark. The course starts from the work trainees already do.

What happens to what our people type into the exercises?

It is stored in the EU, in Ireland. AI requests go through our server-side proxy to Anthropic, which processes them in the US under a data processing agreement and standard contractual clauses. Anthropic does not use API inputs or outputs for training by default.

Can we run it with our own trainers?

Yes, on your own instance. NIBRT runs its own instance with its own trainers as administrators and its own copy of the flagship course.

Does it work without a GMP suite?

Yes. The course runs in an ordinary classroom, and a physical GMP suite is not required.

What does each trainee leave with?

A co-branded certificate, a full course record of their decisions and scores, and an alumni code to keep practising on their own device for 12 months.

How is it different from an e-learning module?

An instructor runs it in a room, trainees work in pairs, and every exercise ends in a decision that is scored and debriefed. E-learning records that someone finished a module. Symerse records the calls they made.

Bring one cohort. See the record they leave with.

Book a demo

A 30-minute call with the engineers who built it. No slides.