One team. A useful path for each role.
Role-based AI training teaches the shared habits of using AI well through the tasks a specific job actually involves. In Tosheo.ai each planned role has its own skill graph, and each learner gets a path through it based on their own results.
A path for your work.
Marketing- 01 / Give AI the right contextCompleted in this example
- 02 / Write a useful first draftCompleted in this example
- 03 / Make it sound like your brandPractice task / in progressYour next move
Rewrite a launch email. Keep the promise specific. Remove the claim you cannot prove.
- 04 / Check before you shareUp next / evidence and judgement
Context
Use a fictional brief, name the audience and set the constraints.
Practice
Make a draft, inspect what changed and try again with clearer instructions.
Judgement
Check evidence, protect information and explain your final decisions.
What is role-based AI training?
Role-based AI training is practice organised around the work a person does, rather than around a tool's menu or a general course. A recruiter practises writing a fair job description and screening questions; a finance colleague practises explaining a variance without inventing its cause. The underlying habits are the same, but the tasks, risks and review decisions are not.
Generic training tends to stop at the prompt. Role training carries on to the decision: is this draft safe to send to this reader, does this figure match the source, does this claim need a second person? That decision is where most of the value and most of the risk of AI at work sits.
Which roles can a manager choose?
The planned preset list is Marketing, Sales, Ops, Support, HR/Recruiting, Finance, Product and General. The manager assigns each invited person to one role, and General suits people whose work does not fit the others, such as an office manager or a founder's assistant.
Custom roles are planned for the Growth plan, for a team whose work falls between presets, such as a client services lead at an agency. Which roles ship first at launch is still an open product decision, so check the current list before you plan a rollout around a specific path.
What is a skill graph?
A skill graph is the defined map of AI-at-work skills for one role. For Marketing, the planned graph covers prompt fundamentals, content drafting, research synthesis, campaign ideation, data summarisation, brand-voice control and AI output checking. The engine picks and orders skills from that map for each learner.
The other rows below are illustrative examples of how graphs can differ by role. The final skill graphs and rubrics for each launch role are still being written.
| Role | Example skills in the graph | Example task |
|---|---|---|
| Marketing | Brand-voice control, campaign ideation, output checking | Rewrite a launch email to match supplied voice examples and remove an unsupported claim |
| Sales | Account research, grounded outreach, call summaries | Draft outreach using only facts from a supplied account note |
| Support | Policy-faithful replies, tone under pressure | Answer an upset customer without promising anything the refund policy does not allow |
| HR/Recruiting | Fair job descriptions, structured interview questions | Turn a hiring manager's notes into criteria and questions that avoid personal assumptions |
| Finance | Variance explanations, figure checking | Explain a change in a supplied table and leave the cause open where the data does not show it |
Can two people in the same role get different paths?
Yes, that is the design goal. The engine is planned to choose and order skills using four inputs: the role, the diagnostic result, the AI tools the team uses and, where provided, company context.
Picture two fictional sales reps at Northfield Software. Priya uses ChatGPT every day but often sends drafts with details the account note does not support, so her path opens with output checking and grounded outreach. Tom has barely used AI and struggles to give it enough context, so his path opens with prompt fundamentals and account research. Both reach the same final assessment by different routes. As they submit work, strong results skip ahead and weak results bring a simpler explanation and a second attempt.
How do the team's AI tools shape the path?
The manager selects the tools the team uses during setup, from ChatGPT, Claude, Gemini, Copilot, Notion AI, Canva AI or other. Tasks are planned to be done in those tools, so the practice happens in the place the work happens.
The habits matter more than the buttons. A path is not a product tutorial for any one tool, and Tosheo.ai does not include subscriptions to them. If you need vendor-specific feature training, the tool provider's own documentation is the better source.
Where do role paths stop?
Role paths start with using AI in the job, and can continue into professional foundations in data science, machine learning, AI and blockchain. Engineers who want hands-on machine learning or coding depth will be better served by a specialist technical course. A path also does not qualify anyone for a regulated activity, such as giving financial or legal advice; it builds the habit of knowing when to hand a task to a qualified person.
At launch, lessons are planned in English only, and customers cannot author their own courses. If your team needs either, a different provider will suit you better for now.
A few good questions.
Which roles will be available at launch?
The planned preset list is Marketing, Sales, Ops, Support, HR/Recruiting, Finance, Product and General. Which of these ship first is still being decided, so check the current list before planning around a specific role.
Can we create a custom role?
Custom roles are planned for the Growth plan. They suit people whose work falls between the presets, such as a client services lead at an agency.
Can one person follow two role paths?
Each invited person is planned to be assigned one role. If someone's work spans two areas, General or a custom role on the Growth plan is likely to fit better than two separate paths.
Can we use role paths to test job candidates?
No. Tosheo.ai trains existing staff. Role paths are training for people already on the team, not a pre-employment assessment.
Keep exploring.
Back to the Tosheo.ai product overview
Take a closer look ↗How the adaptive engine places each learner
Take a closer look ↗Practice scenarios scored against a rubric
Take a closer look ↗AI training for marketers
Take a closer look ↗AI training for sales teams
Take a closer look ↗AI training for HR teams
Take a closer look ↗AI training for finance teams
Take a closer look ↗Compare planned plans, including custom roles
Take a closer look ↗Give your team a path, not another playlist.
Explore the planned role-specific lessons, practical tasks and manager view.