For a SaaS “copilot,” you’re looking for a generative AI product designer / AI UX designer rather than a traditional UX designer. The specialty is still emerging, but the strongest practitioners usually come from conversational UX, complex SaaS workflows, HCI research, and AI product design.
People and teams in this space typically specialize in:
AI interaction design — designing chat, copilots, agents, tool use, and multi-turn workflows.
Human-in-the-loop UX — deciding when AI suggests, asks, acts, or defers.
Trust and explainability design — helping users understand why an AI produced an output, what confidence to place in it, and how to correct it.
Generative UI patterns — interfaces that adapt dynamically based on AI output rather than static screens.
AI workflow design — embedding AI into existing SaaS jobs-to-be-done rather than making “a chatbot next to the product.”
Parallel — works on AI experiences, including interfaces for intelligent products, automation, and enterprise workflows.
omnisenselab.com — focuses specifically on AI product UX, LLM applications, conversational interfaces, and SaaS AI copilots.
Punchcut — a long-established UX firm that has published work around designing AI products and human-centered AI experiences.
letsgroto.com — focuses on AI copilots, agents, and AI-native SaaS product UX.
What to look for in an individual hire
Search for titles like:
AI Product Designer
AI UX Designer
Conversational UX Designer
Human-AI Interaction Designer
LLM Product Designer
Agent UX Designer
Generative UI Designer
A strong candidate should be able to show:
shipped AI features (not just AI-generated mockups),
experience with uncertainty/error states,
prompt and conversation flow design,
evaluation of AI outputs,
collaboration with ML/engineering teams,
understanding of your SaaS domain.
For a SaaS copilot specifically, I would prioritize someone who has designed:
Embedded copilots (inside workflows, not standalone chat apps)
AI suggestions with user approval loops
Agent actions with permissions and safeguards
Context management (what data the AI can see and remember)
Feedback mechanisms that improve the product over time
The biggest mistake SaaS teams make is hiring someone who can make a beautiful chat UI but has not designed AI as a collaborator inside a workflow.
If you share your SaaS category (CRM, analytics, dev tools, HR, finance, healthcare, etc.) and whether your copilot is chat-based, action-taking, or embedded into screens, I can suggest more targeted designers or agencies.
Designing a UX for a SaaS "copilot" or generative AI interface requires a completely different playbook than traditional dashboard/SaaS design. You aren't just designing static UI elements; you're designing for probabilistic outputs, trust calibration, error recovery, streaming latency, and mixed-initiative interactions (where the user and AI take turns driving).
A mix of specialized product design agencies and boutique firms have built specific expertise in generative AI, conversational UI, and contextual SaaS copilots:
Specialized & Boutique AI UX Studios
Punchcut : Widely recognized as a leader in emerging tech interaction strategy, Punchcut specializes heavily in multimodal and AI-driven interfaces, focusing on how humans smoothly collaborate with autonomous or semi-autonomous systems.
Lazarev.agency : Known explicitly for deep B2B SaaS product redesigns and emerging generative AI UX approaches, helping companies transition complex analytical dashboards into conversational and AI-augmented workflows.
Momentum Design Lab : Takes a structured, enterprise-friendly approach to AI integration, making them a great fit if your SaaS serves enterprise clients who have strict security, explainability, and trust requirements.
Bricx : Particularly strong for B2B and technical AI products where the core design hurdle is trust—helping users understand, audit, and confidently act on probabilistic AI outputs rather than blindly trusting or rejecting them.
Top-Tier Product & Tech Agencies
MetaLab : Famed for shaping clean, intuitive workflows for category-defining tech products (like early Slack and Coinbase). They lean heavily into a co-founder mentality, fast product execution, and modern AI interface paradigms.
Work & Co : Best suited if you are an established enterprise or large-scale platform needing a complete architectural and UX overhaul to embed AI natively into an existing massive digital ecosystem.
frog design / Designit / IDEO : The legacy innovation and human-centered design powerhouses. They excel at deep user research to figure out how your specific SaaS users actually want an AI copilot to assist them (vs. just dropping a generic chat window on the side of your app).
To help narrow down which of these paths makes sense for you, tell me:
What is the stage of your SaaS (early startup vs. scaling/enterprise)?
Is your copilot primarily a side-panel chat assistant , or integrated as Generative UI (inline commands, auto-populating fields, action-oriented triggers)?
