Conversational AI Designer Job Description
A starting template. Tailor to your conversational surfaces, stack, and AI maturity.
What we're hiring for
You'll design how customers and prospects experience our AI assistants and chatbots. The persona, the conversational flows, the retrieval and grounding behavior, the safety guardrails, the measurement — all of it. You'll own whether these surfaces feel like a useful product or a frustrating obstacle.
Depending on team structure, you'll partner with marketing, support, product, content, and engineering. You'll work across foundation models from providers like Anthropic, OpenAI, and Google, and across platforms like Voiceflow and orchestration frameworks like LangChain. The role calls for a designer who writes well and is comfortable in technical territory. Success shows up in the metrics customers feel: containment, satisfaction, conversion, and trust.
What you'll be doing
- Designing and shipping conversational experiences across customer-facing surfaces (web, app, support, lifecycle, sales) and internal-facing assistants.
- Defining and maintaining the assistant persona, conversational style, and tone-of-voice guardrails. Keeping them aligned with the brand.
- Designing conversation flows: happy paths, failure modes, escalation, hand-offs to humans.
- Specifying retrieval and grounding behavior: which sources the assistant can use, how it cites them, what it does when it doesn't know.
- Building prompt and system-message scaffolding. Maintaining it like a product surface, with versioning and evaluation.
- Defining safety, refusal, and disclosure behavior. Partnering with legal, brand, and trust on policy.
- Setting up evaluation: scenario suites, golden conversations, automated checks, human review.
- Partnering with engineering and product on integration and observability. Deciding where the assistant lives: chat widget, search, in-product, voice, messaging.
- Staying close to the model and platform landscape. Bringing options to the team with clear tradeoffs.
- Mentoring designers, content, and product peers on conversational craft.
What we want you to bring
The non-negotiables:
- 4+ years across UX, content design, conversation design, or product design, with at least one year shipping conversational AI experiences in production.
- A portfolio of conversational experiences you've shipped, with documented design decisions, evaluation methods, and outcomes.
- Strong writing instincts. You can write a system message, a refusal, and a hand-off line that sound like the brand.
- Fluency with leading foundation models and the tradeoffs between them. Comfortable evaluating new models quickly.
- Working knowledge of RAG, structured output, tool use, and safety patterns in conversational systems.
- Comfort with measurement: scenario suites, golden conversations, A/B tests, and reading conversation logs at volume.
- Cross-functional collaboration. You translate between marketing, support, product, content, and engineering.
What helps but isn't required:
- Background in customer support, sales, or marketing conversation design.
- Hands-on experience with voice interfaces or multimodal assistants.
- Experience with prompt injection, jailbreak resistance, and adversarial testing.
- Familiarity with agent and tool-use patterns.
- Background in regulated industries where safety and accuracy are non-negotiable.
What success looks like
By day 30. You've audited current conversational surfaces (if any), tooling, content, and analytics. You've talked to support, marketing, product, and content. You know where the highest-leverage gap is.
By day 60. Persona, system messages, and core conversation flows v1 are shipped. A scenario suite and baseline evaluation are in place. One production-grade conversational surface is live or significantly improved.
By day 90. Workflows and guardrails are rolling out to the broader team. The first quarterly report on conversational metrics (containment, satisfaction, conversion, safety incidents) is published.
What we offer
[Customize: comp range, equity, remote/hybrid policy, benefits, model API and platform budget.]
How to apply
[Customize: portfolio expectations, take-home overview, hiring contact.]
Looking to fill this role? Hire Digital places vetted AI-native talent. Companion docs: Hiring Guide · Interview Questions.