We’re looking for a Conversational AI UX Designer to define how Loomi (our AI) talks, listens, and guides marketers through their day‑to‑day work in Bloomreach. You’ll design AI‑first experiences across chat, in‑product assistants, and agentic workflows that help marketers create, edit, and understand complex campaigns — without needing to be data scientists or developers.

You’ll work closely with other designers, product managers, and engineers to shape how our agents behave: what they say, what they show, when they ask for more input, and when they quietly handle things on their own. Your work will directly influence how customers trust, understand, and get value from AI in their campaign workflows.

This is a highly collaborative, hands‑on role. You’ll spend a lot of time co‑framing problems with teammates, running working sessions, prototyping conversation flows, and iterating quickly together.

This role offer the possibility of working in one of our Central European offices or from home on a full-time basis.

What You’ll Do

Design conversational and agentic experiences for marketers

• Design and iterate on chat and assistant experiences that help marketers brief campaigns, review proposals, and refine multi‑step workflows.
• Develop and evaluate conversational flows (prompting, turn‑taking, error recovery) across in‑product chat, side panels, and embedded agent UIs.
• Shape Loomi’s voice and tone for marketer workflows: when it should be directive vs. exploratory, when it should show its reasoning, and how it communicates uncertainty.

Treat prompts, context, and reasoning as core design artifacts

• Use prompt and context engineering to steer agent behavior: system prompts, tool selection hints, and contextual grounding from customer data.
• Design for the realities of generative AI — probabilistic outputs, uncertainty, hallucinations, and graceful failure states — with clear guardrails and recovery paths.
• Collaborate with PM and engineering partners to turn prompts, tools, and policies into a coherent UX model that can be evolved over time.

Prototype with real AI behavior, not just static UIs

• Build and test AI‑native prototypes using tools like Claude Code plus working LLM/playground setups (e.g., prompt sandboxes, scripted flows) so teams can experience real behavior, not just mock copy.
• Create prototypes that exercise both happy paths and edge cases (ambiguous prompts, missing data, conflicting inputs) and use those to drive design decisions.

Use data and conversation logs to drive iteration

• Define and track success metrics for agent UX, such as task completion, follow‑up questions needed, error recovery, user edits after agent changes, and perceived trust.
• Work with data and product analytics to mine conversation logs and user behavior for friction points, misunderstandings, and opportunities to improve flows.
• Turn insights from real usage into concrete design changes (updated prompts, new clarification turns, different defaults, revised fallbacks).

Champion quality, safety, and transparency

• Define and help maintain evaluation criteria (evals) for agent responses from a UX lens — clarity, accuracy, safety, tone, and “did this actually help the marketer move forward?”.
• Distinguish between model limitations and UX issues, and advocate for human‑centered guardrails (e.g., when to ask for confirmation, when not to auto‑apply changes, what to keep behind explicit opt‑in).

Design clear explanations and “why” moments so marketers can understand key decisions (segments, timing, splits) without reading a wall of AI output.

What We’re Looking For

Requirements

• Demonstrated experience designing for conversational, assistant, or AI‑powered experiences — chat, copilots, agents, or similar.
• Experience in product/UX design, bonus for complex web‑based or SaaS products.
• Strong collaboration and communication skills, with a track record of:
• Working closely with PMs, engineers, and other designers from problem framing through launch.
• Leading or facilitating design reviews, workshops, or working sessions.
• Explaining design decisions clearly to both technical and non‑technical partners.

• Hands‑on experience with LLM‑driven tools in your own work, for example:
• Using chat assistants or copilots to explore ideas, draft content, or accelerate design/development tasks.
• Prototyping or shipping AI‑powered or assistive product features, ideally where the system suggests or acts on behalf of the user.
• Comfort treating prompts, context, and evaluation criteria as design materials — not just UI text.
• A systems mindset — you think about how patterns scale across surfaces, use cases, and agents, and when it’s appropriate to introduce something new.
• Ability to work independently on end‑to‑end projects while staying tightly aligned with your product trio and design peers.

Nice to Have

• Experience designing for marketing automation, CRM, messaging, or workflow‑driven products, especially for non‑technical users.
• Familiarity with design systems and component libraries (as a contributor or consumer), and comfort extending patterns into conversational/agent surfaces.
• Experience partnering with data and research to analyze conversation logs, define success metrics, and run structured experiments.
• Experience mentoring other designers, or helping teams improve their AI/agent design practices, documentation, or ways of working.

If you’re excited about collaborative problem‑solving, curious about how conversational AI can make marketers’ lives better, and love turning ambiguous agent behavior into clear, trustworthy experiences, we’d love to meet you.

#LI-HO1