Level 2 of three · 24 hours · online, own pace
Your team picks one of five systems — HR, Operations, Logistics, Sales and Marketing, or Customer Service — and builds it fully: customised on the company's own data and documents, with retrieval, guardrails, evaluation and a named owner. One system, built to run and maintained. Not five half-built. Built for the whole company rather than one desk, so more of your people use it, each on their own work. The software is the company's; the machine under it is not — its data and memory sit on a server we provide or recommend, so there is no hardware to buy. Owning the whole thing is Level 3.
L2 is level 2 of the OwnYourAI corporate AI training path. It is 24 hours of online, own-pace learning delivered as 4 modules, and it is the point where your company builds its own AI colleague on a real task.
L2 · 4 modules · not open yet · part of a 72-hour path
Before you book this level
Level 1 is running now. All 4 modules of L2 are written and running in the course platform.
L2 opens when we take the first company onto it. Nothing is charged for L2 today: the price above is what it will cost when it opens, not a bill, and the assessment is free. Tell us you want L2 and we will tell you the day it opens — before you are asked to pay anything.
Plenty of companies start on Level 1 while the next level is being opened up, which is why most take one level at a time.
Session by session
Every session is online learning at the person's own pace, with a scheduled live touchpoint inside it and a trainer present. Each one ends with something your company keeps: 24 hours in total for L2.
| # | Session | The one thing it settles | What your company keeps |
|---|---|---|---|
| 1 | Choosing the right job for an agent | Which workflow is worth automating first, and which must not be | A scored shortlist of three workflows, and the one chosen, with the reason |
| 2 | Mapping the workflow and the data | Where the information lives, who owns it, and what may not be touched | A one-page workflow map plus the data rules for this agent |
| 3 | Writing the agent's brief | What it is told to do, and the three limits it may never cross | The agent's written brief, including the three limits |
| 4 | Building it | A first working version connected to your own information | A first build that answers correctly on three real inputs from your work |
| 5 | Testing it against real work | What breaks, and how you know | A failure log built from your own documents, with the fix for each failure |
| 6 | Guarding it | Who may ask it what, what needs a human, and what gets logged | The guardrail document, plus the corrected rule from the incident drill |
| 7 | Putting it into the daily routine | Who operates it, and what working means on a Monday morning | A one-page operating routine with a named owner and a reviewer |
| 8 | The build assessment | Whether your company actually has a running agent | A live demonstration on fresh inputs, and your company agent record |
The standard every build is held to
A brief that does not name these three is sent back before any building starts. The limits are written by your own staff, in your own words, for your own workflow.
It may do alone
The documented, repeating part of the workflow: reading the same kind of document, drafting from the same template, pulling the same figures together. If a task needs judgement about a person, it is not this.
It must ask a person first
Nothing goes out of the company on its own: not a client letter, not a filing, not a payment instruction. The routine written in Session 7 names who approves, and the approval is recorded.
It never does
Four hard lines, configured as rules rather than left to memory: no sending to a client, no payment, no deletion, and never entering a client's personal data. Session 6 includes an incident drill, so the team has caught it once before it matters.
The assessment
The assessment is delivered live online with a trainer present. It is not a multiple-choice quiz, and it does not test whether someone remembers AI theory: it tests whether they can do the thing on their own desk.
| Part | What the person does | Time | Marked by |
|---|---|---|---|
| A. The agent runs | Runs the agent live on the company workflow, on inputs the assessor supplies fresh at the assessment | 20 min | Trainer |
| B. The failure log | Explains three failures from the log, what caused each one, and what changed as a result | 10 min | Trainer |
| C. The guardrails | Says what the agent may do alone, what needs a human, what it never does, and where the log is | 10 min | Trainer |
| D. The routine | Names the owner, the reviewer, and the three checks that show it is still working | 10 min | Trainer |
A. must succeed on inputs supplied fresh at the assessment: a prepared demonstration does not pass. B., C. and D. must each be answered correctly. Anyone who misses a part is assessed again on that part at no extra cost, and that re-assessment is recorded as such. The assessor is named on the record, with the date.
Per person, kept together and retrievable on its own: the identity record, attendance at each session, progress at each session with the work produced, the assessment result part by part, the artefact that was demonstrated, and the name of the assessor with the date. Everything on that list is written by a named trainer, not by the learner. What we keep, and who can read it is set out on its own page.
A quick look at what you can build
HR, operations, logistics, sales and marketing, and customer service: three screens each, plus why the AI version of each one is different from the software you already pay for.
On illustrative data: no client's system and no client's data appears on those pages.
What this level is for
Before this level
In practice most teams do L1 first, because L1 gives them the ground this level builds on. If your team already has it, the assessment will say so and we will start you here.
After this level
L3 — A Customised Project — Build, Monitor, Grow. Most teams take a short break between levels so the first one settles into daily work.
What your team can do afterwards
These are the skills that travel between companies, which is why the level is written around them. About 70% of every level is that shared skill; about 30% is your own workflows.
| Skill | What the learner hands in |
|---|---|
| Generative AI Model Development and Fine Tuning | The system itself: built on the company's own documents and data, running, and handed over with its settings. |
| Generative AI Model Evaluation | The evaluation sheet: the test questions they wrote, the scores before and after the changes, and the ones it still fails. |
| Prompt Engineering | The system's own instructions, written and versioned, with the reason each rule is there. |
| Workflow Digitalisation | The one repetitive step it replaces, measured before and after. |
| Artificial Intelligence Ethics and Governance | The three limits it works inside, written down, and who may change them. |
| Data Analysis and Interpretation | The company data it reads, mapped and documented, with what is excluded on purpose. |
| Customising an AI Model for a Business System | The finished system, customised on the company's work and run by a staff member without help. |
| What | Detail |
|---|---|
| Level | L2 of three — Build Your Company's Own AI System |
| Hours | 24 hours |
| Delivered as | 4 modules, online and own pace |
| Scheduled touchpoints | Yes — online, booked in advance inside the level |
| Working language | Plain English. No technical background needed. |
| What you keep | The work produced, the written record, and the skill in your own staff. |
A result, how it is measured, when it lands, and what happens if it does not.
L2 — questions
Next step
The assessment tells you whether L2 is the right place to start, and what it would cost for the seats you actually need.
Ask us for a free quotation for the seats you need — nothing is charged before you approve it.