Richard Simms

Principal Product Designer and AI-native product builder. Melbourne, Australia.

I design and build AI-powered products, with a focus on LLM-powered job matching, agentic hiring workflows, semantic layers for product teams, and machine-readable design systems.

I'm currently Principal Product Designer at SEEK, Australia's largest employment marketplace, where I lead design for the GenAI-powered Career Feed. I also build products independently, including Sentiuma (is a portable memory layer that turns your knowledge into a system every AI; Claude. ChatGPT. Cursor. can query.) and Speasy (automated audio content for busy professionals).

Focus Areas

  • LLM-powered job matching

    I led the design of SEEK's GenAI Career Feed, a conversational AI system that helps candidates express job preferences through natural language rather than filters. The system captures 21.4 preference entities per chat session (vs. 2.3 per traditional search interaction) and collects 10× more candidate intent signal than conventional search. This is now the foundation of SEEK's AI recommendations strategy.

  • Agentic career workflows

    I design agentic experiences systems where AI takes sequences of actions on behalf of users, not just answers single questions. In job-matching, this means understanding candidate intent across multiple turns, surfacing high fit roles, and closing the loop between candidate preference and recommendation quality.

  • Your knowledge should not live inside chat windows.

    Sentiuma is my answer to that problem. A portable memory layer built on plain markdown, queryable by any AI, owned entirely by you. No vendor lock-in. No proprietary format. Just your knowledge, made legible to models.

  • Machine-readable design systems and workflows

    I work at the intersection of design and engineering using Markdown-native, version-controlled workflows so design artefacts are usable by both humans and AI. The design system becomes a machine-readable protocol, not a static spec.

  • Prompt design for conversational UX

    I design the UX layer of LLM-powered products by working directly in model playgrounds (Anthropic Workbench, Amazon Bedrock, OpenAI Playground) not just in Figma. The best conversational experiences are shaped in the prompt, through synthetic data testing, multi-turn simulation, and edge case generation.

Notable Projects

  • GenAI Career Feed — SEEK (FY25)

    Led design from zero to launch for SEEK's first GenAI-powered job recommendation feature. Ran a 3-week beta with 45% of Australian logged-in traffic.

    • 21.4 preference entities per chat (vs. 2.3 per search — 9× more signal)
    • 60% completion rate for preference-setting conversations
    • ~12% adoption overall (~20% for active recommendations users)
    • Experiment insights now power SEEK's recommendations strategy

  • Sentiuma

    Sentiuma is a portable memory layer that turns your knowledge into a system every AI; Claude. ChatGPT. Cursor. can query. One brain. Every AI. Sentiuma is not a note-taking app. It is the layer that makes your existing notes useful to AI.

  • Speasy

    Email-to-audio conversion app. Transforms saved articles and newsletters into a personal podcast feed. Built as a no-code/low-code MVP using AI APIs; expanded into a full-stack product with Supabase, custom audio processing, and RSS feed generation.

Point of View

The best AI products aren't designed in Figma. They're shaped in the prompt.

Most designers are still designing with gloves on mocking up conversational interfaces without ever stress-testing the actual LLM behaviour. I prototype directly in model playgrounds: feeding structured production inputs, generating synthetic variants, testing multi-turn flows. This is where tone, nuance, and helpfulness actually emerge.

Markdown is the new design system for AI-collaborative teams.

Traditional product tools like Miro, PowerPoint, Confluence that produce “context islands” that LLMs can't navigate. Markdown files in version-controlled repositories are the universal protocol for human-AI collaboration. Product strategy as code. Discovery artefacts that the whole team, plus AI, can reason over.

Good design should pay — for the user, the business, and the designer.

My decade mission is to help designers prove the business impact of their work. Not through vanity metrics, but through documented chains of impact: insight → design decision → product metric → business outcome.

Background

  • Principal Product Designer, SEEK (Melbourne) — leading design for AI-powered job discovery
  • 10+ years in product design across employment technology, fintech, and SaaS
  • Strengths: Design strategy, AI product design, skills-based matching systems, prompt design, design-to-code workflows
  • Builder: Sentiuma, Speasy, Design ROI Dashboard
  • Based in Melbourne, Australia

Writing and Talks

I write about AI product design, LLM-powered workflows, and the business impact of design at rsimms.com and on Substack.

Topics I write about:

  • Designing agentic job-matching workflows
  • Semantic layers for AI-native product teams
  • Prompt design and conversational UX
  • Spec-driven discovery with AI
  • Measuring design ROI

All writing →

Contact

Richard Simms is a product designer and builder based in Melbourne, Australia. He specialises in LLM-powered job matching, agentic product design, and machine-readable design systems. He is Principal Product Designer at SEEK and founder of Sentiuma.