A guide to AI tools and workflows for product teams

Perplexity functions as an AI-powered search engine designed to simplify technical research. It synthesises information from trusted sources into clear, concise responses, making it ideal for understanding industry-specific terms, acronyms and client nuances.

• Efficient for technical and industry-specific research.

• Provides reliable citations for its outputs.

• Simplifies complex jargon for better comprehension.

• Limited in generating creative outputs.

• Primarily focused on factual retrieval rather than ideation.

Market research and project insights. Perplexity can aid teams in pulling detailed reports and breaking down technical terms during project kick-offs.

Leverage Perplexity to gather accurate background information and validate research findings before incorporating them into broader design workflows.

Claude is a versatile AI assistant excelling in research, brainstorming and summarisation tasks. It can generate market analyses, user personas, user stories and visual data representations like flowcharts. Its reasoning capabilities allow it to create low to medium-fidelity wireframes and structured layouts for design.

• Strong at reasoning and creating structured outputs.

• Can process user stories and context to generate user-flow diagrams.

• Supports Mermaid JS for visual representations.

• Versatile across multiple product design tasks.

• Limited high-fidelity design capabilities.

• Requires iterative prompts for detailed results.

Early-stage ideation and research. Claude helps designers create user personas and visualise user flows, offering clear diagrams and structured ideas to build a foundation for projects.

Use Claude for initial project planning, creating personas and developing rough wireframes. Pair with more design-specific tools for polished outputs.

NotebookLM is well suited for organising and synthesising data from multiple sources. It allows users to structure information into separate notebooks, offering clear outcomes and summaries for further use.

• Excellent for collating and managing large datasets.

• Can accept multiple input types, including documents, MP3s and even YouTube links.

• Outputs are concise and context-aware.

• Provides suggestions based on uploaded project data.

• Limited generative design capabilities.

• Requires thorough initial data inputs for the best results.

Organising research data and creating briefing documents. NotebookLM is invaluable for consolidating workshop outputs and generating cohesive project overviews.

Use NotebookLM to manage multi-source data during the research phase and generate summaries for team alignment.

ChatGPT is a versatile chatbot ideal for brainstorming, generating scripts and handling translation or analysis tasks. It’s a strong all-rounder for ideation and refining product documentation.

• Intuitive interface and general-purpose functionality.

• Adaptable to various tasks, from planning to user story creation and interview script writing.

• Outputs may require additional refinement for precision.

• Less specialised for technical or visual design.

Brainstorming and stakeholder engagement. ChatGPT is an effective starting point for drafting interview scripts, exploring user needs and generating creative ideas.

Use ChatGPT as a brainstorming tool, and a user story and script generator. Refine its outputs with domain expertise for actionable results.

v0 is an AI-driven UI design system that generates React code compatible with frameworks like Tailwind CSS. It transforms text prompts into functional UI components and supports iterative refinement.

• Generates frontend code alongside UI designs.

• Encourages iterative improvement through prompts.

• Provides rationale for design decisions.

• Includes integrated Figma design capabilities.

• Limited flexibility in styling beyond initial constraints.

• Best suited for developers comfortable with React.

Rapid prototyping of web interfaces. v0 is a practical tool for converting wireframes into working prototypes with minimal effort.

Use v0 for quick prototyping and ideation. Combine with a designer’s manual tweaks for aesthetic polish.

Bolt specialises in high-fidelity design and frontend development. It creates polished, branded UI components and supports visually engaging layouts with animated interactions.

• High-quality, visually appealing designs.

• Capable of integrating imagery and animations.

• Good for inspiring layout and styling choices.

• Requires specific and iterative prompts for best results.

High-fidelity design. Bolt is ideal for creating visually compelling final designs and showcasing potential interfaces to stakeholders.

Employ Bolt for polished presentations and stakeholder reviews. Combine its outputs with functional prototypes for end-to-end design clarity.

Credit: This image was created using ChatGPT (OpenAI)

Scenario: A product designer needs to understand an industry or market that they are unfamiliar with.

• Tools: Perplexity, Claude.

• How AI helps: Use Perplexity to break down technical terms and gather market insights. Employ Claude to summarise findings into key trends, challenges and opportunities.

• Outcome: Designers quickly build foundational knowledge for strategic decision-making.

“Using current industry reports, summarise the top three trends, challenges, and opportunities in the [specific industry] sector. Focus on insights relevant to product design.”

Screenshot: Perplexity.AI

Scenario: A team is defining potential users for a product.

• Tools: Claude, NotebookLM.

• How AI helps: Claude generates detailed personas, including demographics and goals. NotebookLM organises the data for ongoing reference.

• Outcome: Comprehensive, reusable personas for design alignment.

“Create detailed user personas for a [type of product]. Include demographics, goals, behaviours, pain points, and technology familiarity.”

Scenario: Designers need to brainstorm features and visualise user flows.

• Tools: Claude, v0.

• How AI helps: Claude creates user-flow diagrams and structured wireframes. v0 transforms these into functional prototypes.

• Outcome: Efficient ideation with tangible outputs.

“Based on the following user stories, generate user-flow diagrams for [specific feature]. Include key decision points and alternate paths.”

Scenario: Product managers conduct stakeholder interviews.

• Tools: ChatGPT, NotebookLM.

• How AI Helps: ChatGPT drafts tailored interview scripts, while NotebookLM organises feedback into actionable insights.

• Outcome: Improved understanding of stakeholder needs and expectations.

“Create an interview script for [stakeholder type]. Include open-ended questions about [specific product aspects], pain points, and expectations.”

Screenshot: ChatGPT (OpenAI)

Scenario: A team needs to create detailed functional prototypes for showcasing to stakeholders.

• Tools: Bolt, v0.

• How AI helps: Bolt and v0 both generate functional frontend code with an aesthetically pleasing visual style for review and refinement.

• Outcome: Ready-to-use prototypes for stakeholder buy-in and developer alignment.

“Design a high-fidelity interface for [specific page/feature]. Emphasise [design elements] and align with [branding/style].”

Credit: This image was created using ChatGPT (OpenAI)

Let’s consider an example scenario; A banking organisation is interested in creating an investment platform app for its customers. After initial discussions and agreements, an Ignite workshop is planned by Nearform. Our discovery team is assembled and in the days leading up to the workshop, they familiarise themselves with, and bolster their knowledge of the client and the industry. The team make extensive use of Perplexity, with its strong technical research capabilities, fact-checking and jargon simplification.

Screenshots: ChatGPT (OpenAI), Perplexity.ai, Claude.ai (Anthropic), NotebookLM (Google)

The team then centralises their research using NotebookLM, importing collected data and linking relevant documentation. Through carefully crafted prompts, they generate a concise briefing document that captures the essence of the client's business, industry landscape and project requirements. The AI tool assists in identifying potential opportunities and exploration avenues, providing a strategic overview that prepares the team for meaningful discussions.

The team use the innovative ‘Audio Overview’ feature, creating a podcast-style summary that allows team members to absorb key insights even when away from their desks - maximising knowledge retention and accessibility.

Utilising Claude, the team begins shaping user personas, bringing critical empathy and user-centered thinking to the upcoming workshop. By leveraging the preparatory documentation, they can develop nuanced representations of potential platform users, ensuring the product design remains deeply connected to real-world needs.

During the workshop, the team enters with much confidence. Their AI-assisted preparation allows for more focused, insightful conversations. As discussions unfold and sticky notes accumulate, NotebookLM is ready to organise and summarise the captured insights, creating comprehensive overviews and suggesting potential areas for deeper exploration.

Post-workshop, the team synthesises the collected data, documentation and workshop outcomes. Then, they rapidly generate a concise product brief, followed by a high-level set of epics and user stories. Claude then assists the product design team in visualising product flows, facilitating stakeholder alignment and laying the groundwork for subsequent ideation and prototyping phases.

This AI-enhanced workflow transforms traditional product discovery, enabling teams to work smarter, move faster and have higher-level conversations with stakeholders. By intelligently augmenting human creativity and insight, these tools are reshaping how product teams approach complex challenges.

When using AI tools in design workflows, there are several important ethical considerations to keep in mind. Firstly, AI systems can inherit and amplify societal biases. Product managers and designers need to be vigilant about monitoring for bias in the outputs and taking steps to mitigate it.

There are also concerns about the transparency and explainability of AI-generated designs, as well as issues of intellectual property and data privacy. Designers must ensure they have clear policies in place and obtain proper consent when using customer data.

Additionally, AI tools could have unintended consequences, like the merging of opposing design styles. So, designers should carefully evaluate the generated outcomes, and be prepared to step in to manually adjust with the human touch.

Finally, maintaining and updating AI-powered design workflows can be challenging, and teams should stay up-to-date on the latest developments to ensure long-term sustainability.

By proactively addressing these ethical considerations and challenges, product teams can harness the power of AI while upholding their professional responsibilities to solve complex problems and generate impactful outputs more efficiently than ever before.

Product designers should strategically combine tools based on project needs, leveraging external data and meeting and workshop outcomes to enhance prompts.

AI tools like NotebookLM and Perplexity support research, Claude aids synthesis and ideation, while Bolt/v0 accelerate prototyping. Using tools in combination, discovering effective sequences of prompts for different tools and scenarios can help to maximise output and efficiency.

Maintaining iterative workflows ensures AI outputs align with project goals and AI should augment - not replace - human expertise. Always validate results and refine with the human touch.