Browse categories
Explore
Fiverr Pro
English
$
USD
I will design your ai chatbot persona with voice, tone, and scripts
Mandy Hathaway
About this gig
Your AI agent needs a personality that holds up under pressure. Not a tagline and a wave emoji. A real character with consistent tone, honest handling of uncertainty, and conversation flows that build trust instead of selling.
I design AI personas at the specification level: identity and voice architecture, tone guidelines with concrete examples, conversation scripts stage by stage, and behavioral guidelines your engineers can build from. GDPR and EU AI Act-aware by default.
Recent project: a 13-document persona package for a UK AI talent marketplace. System prompts, archetype classification, cross-surface coherence across chat and voice, red team testing with mitigations. 85% production-ready on delivery.
Background: MA in Ethical Technology and AI, 30+ years in tech, AI ethics consultant with a public writing practice. I write the specs that engineers want and the voice that customers actually trust.
Message me before ordering so we can scope your use case. Every persona is custom to your product, audience, and deployment channel.
Get to know Mandy Hathaway
Mandy Hathaway
Technical Writer and AI Ethics Consultant
- FromUnited States
- Member sinceMar 2026
- Avg. response time9 hours
- Last delivery1 day
Languages
English
Technical writer and AI ethics consultant. 30 years in tech, MA in Ethical Technology and AI. I design AI personas and conversation architectures for products that need a voice users actually trust, and I write technical documentation that reflects how these systems really work. Clients: Zeki Research (AI career platform for women in tech, 13-document persona spec for a conversational agent across chat and voice) and eDynamic Learning, Pearson (AI curriculum for high school CS). Founder of aimissionready.org. I'm honest about what AI can and can't do, and I build specs engineers want to ship.
My Portfolio
FAQ
Do you build or code the chatbot?
No. I design the persona, tone, and conversation architecture at the specification level. The deliverables are documents your engineering team uses to build or configure the bot. If you need development, I can refer you or quote it as a separate engagement.
Which LLM or platform does this work with?
The specs are platform-agnostic by design. I've worked with Claude, GPT, and custom API setups, and the documents are written so your engineering team can implement on any modern LLM stack (OpenAI, Anthropic, open-source, Rasa, Dialogflow, custom).
What do I need to send you to start?
A few things help: what the bot does, who uses it, where it runs (chat, voice, both), any brand voice docs, and example conversations you want it to handle. Don't have all of this? Send what you have and we'll fill the gaps in the kickoff call. Most clients send 1-2 docs.
Can you work with an existing bot that already has a persona?
Yes. I can audit what you have, identify consistency gaps, and rewrite at the component level (tone guide, scripts, system prompts) rather than starting from scratch. Message me first so we can scope it accurately.
Do you handle ethics and compliance?
Yes. Every persona I deliver aligns with major responsible AI frameworks (NIST AI RMF in the US, EU AI Act and GDPR in Europe) and adapts to your deployment jurisdiction. I hold an MA in Ethical Technology and AI, and the Premium tier includes a dedicated ethics and compliance layer with a red team
