I design with Claude more than Figma now

For years, financial professionals have relied on a toolkit of spreadsheets, complex formulas, and dedicated design software like Figma to visualize and communicate complex data. While these tools remain essential, a new player is rapidly changing the game: Artificial Intelligence, specifically large language models (LLMs) like Claude. I've personally found myself leaning on Claude more and more, and frankly, more than Figma, for a surprising range of financial tasks. This isn’t about replacing traditional tools entirely; it's about augmenting them and unlocking new levels of efficiency and insight.

§The Traditional Workflow: Figma and Its Limitations in Finance

Figma is a fantastic tool. It’s a collaborative, cloud-based design platform excellent for creating visually appealing presentations, dashboards, and user interfaces. In finance, we've used it for:

• Building investor decks: Showcasing performance metrics and future projections.
• Designing client reports: Presenting complex financial data in an easy-to-understand format.
• Prototyping financial applications: Mocking up UIs for internal tools or client-facing platforms.
• Creating infographics: Simplifying complex concepts with visual representations.

However, Figma has inherent limitations when applied to the dynamic nature of finance. It's largely static. Any changes to underlying data require manual updates across multiple designs. This is time-consuming, prone to errors, and doesn't readily accommodate iterative "what-if" scenarios – a cornerstone of financial modeling. The process is often:

1. Data analysis in a spreadsheet (Excel, Google Sheets).
2. Manual transfer of data to Figma.
3. Visual design and chart creation within Figma.
4. If data changes…repeat from step 1.

This creates a disconnect between the analysis and the presentation, increasing the risk of inconsistencies. It's a fundamentally reactive process.

§Enter Claude: The AI-Powered Design Assistant

Claude, developed by Anthropic, is an LLM that excels at understanding and generating natural language. But its capabilities extend far beyond simple text generation. Claude can:

• Understand complex financial concepts: It’s been trained on a massive dataset, including financial reports, academic papers, and market data.
• Generate code: Python, R, even snippets of JavaScript for interactive charts.
• Interpret data: Provide summaries, identify trends, and draw conclusions from financial data sets (when fed the data).
• Create visuals from text prompts: This is the key shift. Instead of manually dragging and dropping elements in Figma, you describe the visualization you want, and Claude can generate it (often as code that you can then embed elsewhere).
• Iterate rapidly: “Make the line chart bolder,” “Change the color scheme to match our brand guidelines,” “Show a five-year projection instead of three” – Claude can handle these requests instantly.

§How I Use Claude More Than Figma Now: Real-World Examples

Here's a breakdown of how I've integrated Claude into my financial workflow, and why it’s often my go-to over Figma:

• Rapid Prototyping of Investment Scenarios: Instead of building static mockups in Figma, I now describe the investment scenario to Claude: "Create a table showing projected returns for a portfolio with 60% stocks, 30% bonds, and 10% real estate, using historical data from the last 20 years. Include columns for average return, standard deviation, and Sharpe ratio.” Claude can output this as a markdown table, or even Python code to generate a chart using libraries like Matplotlib or Seaborn. I can then quickly tweak the parameters and regenerate the visualization.
• Generating Client-Ready Reports: I feed Claude the core financial data and a prompt like, “Create a one-page summary report for a client, highlighting key performance indicators (KPIs) and explaining our investment strategy in plain language. Include a chart showing portfolio allocation.” Claude produces a well-structured report draft that I can then refine and customize. This drastically reduces the time spent on report creation.
• Building Interactive Financial Dashboards (with Code): While Claude doesn’t directly create fully interactive dashboards like you’d build in Tableau or Power BI, it generates the underlying code (Python with libraries like Plotly or Dash) that I can then use to build these dashboards. This gives me greater flexibility and control.
• Explaining Complex Financial Concepts to Non-Technical Audiences: I can ask Claude to “Explain the concept of discounted cash flow analysis in simple terms, suitable for a client with no financial background.” It provides clear, concise explanations that I can incorporate into presentations or reports.
• A/B Testing Visualizations: Need to see if a bar chart or a line graph better conveys a specific data point? I can ask Claude to generate both, quickly compare them, and choose the most effective option.

§Claude 3: A Step Change in Capabilities

The release of Claude 3 (Opus, Sonnet, and Haiku) represents a significant leap forward. Its enhanced reasoning abilities, faster response times, and improved ability to handle complex instructions make it even more valuable for financial professionals.

• Opus: The most powerful model, ideal for complex financial modeling and analysis requiring nuanced reasoning.
• Sonnet: A balance of speed and intelligence, perfect for generating reports and summarizing financial data.
• Haiku: The fastest and most cost-effective model, great for quick tasks like data exploration and generating simple visualizations.

The improved context window of Claude 3 also allows me to feed it larger datasets and more detailed instructions, leading to more accurate and insightful results. I've found Opus particularly good at debugging complex financial calculations embedded in Python code.

§The Limitations & The Future of AI-Driven Financial Design

§Claude isn't a silver bullet. There are limitations:

• Data Security: Sharing sensitive financial data with any AI platform requires careful consideration of security and privacy. Ensure the platform has robust data protection measures in place. (I always anonymize data where possible.)
• Accuracy Verification: Claude can sometimes hallucinate or provide inaccurate information. It’s crucial to always verify its outputs, especially when dealing with financial data. Treat it as an assistant, not an authority.
• Code Debugging: While Claude can generate code, debugging it still often requires a solid understanding of programming.
• Lack of True "Design" Sense: Claude generates visuals based on instructions, but it doesn't have the artistic flair or design intuition of a skilled graphic designer using Figma. It's excellent for clarity and function, less so for high-end aesthetics.

§However, the future looks incredibly promising. We can expect:

• Improved Data Integration: Seamless integration with financial data providers and platforms.
• More Sophisticated Visualizations: AI models capable of generating truly stunning and informative visuals.
• Automated Report Generation: Fully automated report generation based on real-time data and pre-defined templates.
• Personalized Financial Advice: AI-powered tools that can provide personalized investment recommendations.

§Is Figma Obsolete?

No, not at all. Figma remains the gold standard for pixel-perfect design and collaborative editing. However, for the initial stages of financial modeling, data visualization, and report prototyping, Claude offers a speed and flexibility that Figma simply can't match. I see it as a complementary tool – Claude for rapid iteration and code generation, Figma for final polish and presentation. My workflow now centers around starting with Claude and finishing with Figma where refined design is critical. I'm currently exploring plugins like https://example.com/ for Figma that help bridge the gap between AI-generated content and design customization. And I’m using https://example.com/ to keep my hardware running smoothly while I explore these new AI tools.

This shift from manual design to AI-assisted workflows isn't just about saving time. It's about unlocking new levels of insight and making better financial decisions.

Disclaimer: I may earn a commission if you purchase products through the affiliate links in this article. This does not affect my recommendations or opinions.