AI Persona Generator

The AI Persona Generator designs detailed, repeatable personas for role-play or assistant use inside any LLM workflow. You can start from the pre-filled example — Maya Chen, a Senior Product Designer — or hit Randomize persona to get a fresh name, role, background, speaking style, and two to four matching traits. The Identity card collects a persona name, a role or profession, an expertise slider from 1 to 10 with live labels running from Beginner through Expert, a set of eight trait checkboxes, and free-form background and speaking-style textareas, all reflected instantly in a row of status badges. Two tabs render the results: the Profile tab produces a formatted Markdown profile with role, expertise level, traits, background, and speaking style, while the System prompt tab composes a ready-to-paste instruction that casts the model as the persona, folds in the background and style, and closes with a stay-in-character directive. Both outputs can be downloaded as Markdown files. Because everything is generated locally from your inputs, the tool suits product demos, support agents that need a consistent tone, game characters, and any scenario where a stable, documented persona beats an improvised one.

1. 1

   Click "Randomize persona" for a fresh combination, or keep the default Maya Chen example and edit each field.

2. 2

   Fill the Persona name and Role / profession inputs in the Identity card — empty names render as "Unnamed".

3. 3

   Drag the Expertise level slider and watch the hint update its label, for example "8/10 · Advanced".

4. 4

   Tick one or more of the eight trait checkboxes; the counter above the grid shows how many are selected.

5. 5

   Write free-form Background and Speaking style text, then open the System prompt tab and download it as persona-system-prompt.md.

• Casting a consistent tone for a customer-support chatbot so every reply matches the brand voice.
• Creating characters for prompt-based games or interactive fiction with a documented identity.
• Preparing a mock user persona with a distinct background for UX interviews or usability scripts.
• Generating a documented buyer profile for sales teams to role-play objections against.
• Versioning personas as Markdown files so prompt changes stay reviewable in a repository.
• Building study aids in which each persona explains a topic from a different expertise level and perspective.

• Keep the background short and concrete — one or two sentences anchor the model far better than a long résumé.
• Limit traits to two or three for clarity; the sample picks three, and contradicting traits like humorous plus serious dilute the character.
• Match the Expertise level to the role, since a "Staff Engineer" rated Lv 2 reads as inconsistent.
• Name and frame the role first, then tailor the speaking style to the channel, terse and direct for chat for example.
• Download both outputs so the system prompt and its documentation stay in sync as you iterate.
• Leaving traits unselected still produces a workable "neutral and balanced" prompt when none fit.

• Randomize repeatedly to break writer's block, then edit the persona that feels closest to your goal.
• Copy the System prompt output straight into the "system" role of any chat API call.
• Keep traits to two or three so the persona stays legible across long conversations.
• Store both downloads together so the prompt and its documentation never drift apart.