How to Use Perplexity AI: Build 3 Projects Step-by-Step

What is Perplexity AI

Perplexity AI is an AI-powered search engine that provides real-time, cited answers to user questions instead of returning lists of links like traditional search engines. This tutorial demonstrates how to build three practical projects using Perplexity AI: an academic literature review with proper citations, a market research report with an interactive dashboard, and a real-time news aggregator. Each project includes step-by-step instructions, specific prompts to use, and guidance on choosing between research mode and labs mode. No coding experience needed.

How Perplexity AI works

Perplexity AI combines natural language processing with real-time web search to deliver comprehensive, cited responses. The platform operates through three distinct modes designed for different use cases: Search Mode delivers quick answers to specific questions using Perplexity’s fast AI model. This free mode returns formatted answers instead of link lists, making it ideal for rapid fact-checking and general queries.

1. Research Mode performs comprehensive research by analyzing hundreds of sources and generates detailed reports with citations in 3-5 minutes. Perplexity research mode excels at academic literature reviews, competitive analysis, and market research where comprehensive sourcing is critical.
2. Labs mode transforms research into interactive applications, building complete projects including dashboards, visualizations, and web apps. Perplexity Labs generates downloadable HTML files and deployable applications in 10-30 minutes, eliminating the need for coding knowledge.

What makes Perplexity different

Perplexity’s architecture combines multiple AI models (GPT-4, Claude, and proprietary models) with live web indexing. Unlike ChatGPT, which generates responses from training data, Perplexity actively searches current web content and provides verifiable source citations for every claim. Compared to Google Search, Perplexity delivers structured, analyzed answers rather than requiring manual research across multiple websites.

Perplexity plans & pricing

Perplexity offers multiple subscription tiers with varying capabilities:

Requirements for this tutorial:

• Perplexity Pro subscription
• Web browser (Chrome, Firefox, Safari, or Edge)
• 10-15 minutes per project
• Basic understanding of research workflows

With these fundamentals established, the following sections demonstrate building three practical projects that showcase Perplexity’s research and creation capabilities.

Build an academic literature review with Perplexity research mode

Academic literature reviews require comprehensive research across multiple sources with proper citations. Research mode handles this by analyzing scholarly publications and generating structured reviews in minutes.

Step 1: Setting up research mode

Navigate to perplexity.ai and look at the icons at the bottom of the search box. Click the icon that activates research mode (this appears as one of the mode options alongside the standard search). Perplexity research mode performs deeper analysis across multiple sources compared to the quick search mode.

Before entering the prompt, enable Academic Focus by clicking the “Focus” button in the search interface. Select “Academic” from the dropdown options. This directs searches to scholarly databases, peer-reviewed journals, and academic publications rather than general web content.

Step 2: Creating the literature review

Enter a detailed prompt specifying the research topic, scope, and structure requirements. Here’s an example prompt for a transformer architecture review:

Create a comprehensive literature review on transformer architecture in natural language processing. Include 15-20 citations from peer-reviewed papers published between 2017-2024. Structure the review with the following sections: foundational concepts, key innovations, architectural variations, applications, limitations, and future directions. Focus on seminal papers and highly-cited recent advances. Include full citations with DOI links. 

Copy to clipboard
Copy to clipboard

Want to write better prompts? Check out Codecademy’s free Learn Prompt Engineering course to learn techniques that get better results from AI tools.

Click submit and Perplexity research mode begins processing. The interface displays real-time progress as the system searches databases, reads papers, and synthesizes information. This process takes 3-5 minutes depending on topic complexity.

Research mode generates a 2,000-3,000 word literature review with proper academic structure. The output includes:

• Section headings matching the requested structure
• Inline citations formatted as superscript numbers
• A complete references section with full bibliographic details
• DOI links for each cited paper
• Synthesis of findings across multiple sources rather than summarizing individual papers

The citations come from verified academic sources including arXiv, PubMed, IEEE Xplore, ACM Digital Library, and Google Scholar. Each citation includes authors, publication year, journal or conference name, and DOI when available.

Perplexity research mode saves 4-6 hours compared to manual literature review processes. The generated review provides a solid foundation that can be refined and expanded based on specific research requirements.

The next project demonstrates combining research mode with labs mode to create market analysis reports with interactive visualizations.

Create a market research report with Perplexity AI

Market research reports require both comprehensive analysis and visual data presentation. This project uses a two-phase approach: Research mode generates the analytical report, then labs mode creates an interactive dashboard with downloadable charts.

Step 1: Generating the research report

Start by activating research mode using the mode selector icons at the bottom of the search box. Unlike the academic review, market research uses Web Focus instead of Academic Focus. Click “Focus” and ensure “Web” is selected to access business databases, industry reports, and market analysis sources.

Enter a detailed market research prompt. This example analyzes the AI coding assistant market:

Conduct a comprehensive market analysis of AI coding assistants including GitHub Copilot, Cursor, Replit AI, and Amazon CodeWhisperer. Include: market size and growth projections (2023-2028), user adoption rates, pricing models, key features comparison, competitive positioning, major funding rounds, customer segments, and market challenges. Provide data from industry reports, company announcements, and market research firms. Include specific revenue figures and user statistics where available. 

Copy to clipboard
Copy to clipboard

Research mode processes the request in 4-6 minutes. The system searches business intelligence sources, reads company reports, analyzes market data, and generates a structured report.

Each data point includes citations to source material such as industry reports from Gartner, IDC, or Statista, company press releases, and financial disclosures. The report provides specific numbers rather than vague estimates.

Step 2: Creating interactive dashboards with labs

After reviewing the research report, switch to labs mode by clicking the lightbulb icon at the bottom of the search box. This takes longer than Research but generates complete interactive applications.

Copy the entire research report content and paste it into labs along with a dashboard creation prompt:

I'm providing a comprehensive market research report on AI coding assistants. Create an interactive dashboard that visualizes the key data points from this report. The dashboard should include: 

1. Market size comparison chart showing growth from 2023 to 2028 
2. User adoption statistics for GitHub Copilot (26M users), Cursor (1M users), Replit (40M users), and Codeium (700K users) 
3. Revenue comparison showing Cursor ($200M ARR), Replit ($150M ARR), and GitHub Copilot subscription numbers 
4. Pricing comparison table for all platforms (individual, pro, enterprise tiers) 
5. Funding comparison chart showing Cursor ($160M total), Replit ($250M Series C), and Codeium ($150M Series C) 
6. Feature matrix comparing code completion, agent mode, chat interface, and enterprise features across platforms 

Make the dashboard interactive with filters for time period and company selection. Use a professional color scheme with dark mode support. Include data export functionality for all charts. 

[Paste full research report here] 

Copy to clipboard
Copy to clipboard

Labs mode begins automated project creation, taking 10-15 minutes. The interface shows progress through multiple tasks: processing the research data, generating visualizations, writing code, and deploying the dashboard.

Step 3: Test the dashboard components and assets

Labs generates an interactive dashboard accessible through the Apps tab. The dashboard runs as a standalone web application with:

• Interactive charts using Chart.js or similar libraries
• Dropdown filters for company and time period selection
• Hover tooltips showing detailed data points
• Responsive design working on desktop and mobile
• Dark mode toggle for different viewing preferences
• Export functionality for individual charts as PNG files
• CSV export for underlying data tables

The dashboard includes built-in data export features. Click export buttons on individual charts to download them as PNG images, or use the CSV export buttons to download pricing data, revenue comparisons, and feature matrices as spreadsheet-compatible files.

For online sharing, click the “Share” button in the top-right corner of the deployed dashboard. Two sharing options are available:

• Private: Only the author can view the dashboard. This setting keeps the project restricted to the Perplexity account that created it.
• Anyone with the link: Generates a shareable URL accessible to anyone without requiring Perplexity login. The link remains active and provides full dashboard interactivity including filters, chart exports, and data downloads.

The shareable link works independently of the Perplexity platform, allowing team members or stakeholders to access the dashboard through any browser. Changes to the original project do not update the shared link automatically.

The final project demonstrates building real-time monitoring dashboards that update automatically with current information.

Build a news dashboard with Perplexity labs

News dashboards aggregate current information and present it through interactive visualizations. Labs creates these dashboards by pulling real-time web data and generating complete web applications with filtering capabilities.

Step 1: Creating a space for organization(Optional)

Before building the dashboard, consider creating a Space to organize this project. Spaces help keep related work together but are not required for labs functionality.

Click “Spaces” in the left sidebar, then click “Create a space”. Enter a title like “AI News Monitoring” and optionally add a description. This step is entirely optional, labs works the same whether created in a Space or directly from the home screen.

Step 2: Accessing labs mode and creating the dashboard

Click the lightbulb icon at the bottom of the search box to activate labs mode. This switches from standard search or research mode to the project creation environment.

Enter a comprehensive prompt that specifies the dashboard structure, data sources, and interactive features:

Create an interactive news dashboard that tracks and displays the latest developments in AI regulation globally. The dashboard should include: 

1. Timeline view showing major AI regulation events from 2024-2025 
2. Geographic map highlighting countries with active AI legislation 
3. News feed with the 15 most recent articles from reputable sources (Reuters, Bloomberg, government sites) 
4. Category filters for: privacy laws, safety standards, employment regulations, copyright issues 
5. Search functionality to find specific regulations or countries 
6. Dark mode toggle 
7. Source credibility indicators for each news item 

Pull data from official government sites, major news outlets, and regulatory body announcements. Include article headlines, publication dates, source names, and brief summaries. Make the interface clean and professional with responsive design. 

Copy to clipboard
Copy to clipboard

Labs begins processing, which takes 10-20 minutes depending on complexity. The interface displays progress through multiple tasks: researching current AI regulation news, structuring data, generating visualizations, writing application code, and deploying the dashboard.

Step 3: Testing dashboard features

The Apps tab displays the deployed dashboard with three main views:

• News Feed shows 15 recent articles with headlines, sources, credibility ratings (star indicators), and category tags. Click articles to expand full content.
• Timeline presents regulations chronologically with color-coded dots by region. Click timeline cards for detailed descriptions.
• Regulatory Map displays countries color-coded by status: dark green (comprehensive), light green (in-progress), yellow (guidelines), orange (proposed). Click countries to view specific laws.
• Sidebar filters narrow content by category, with item counts displayed. Search functionality in the top bar filters across all views. Dark mode toggle switches themes via the sun/moon icon.

The dashboard displays data from creation time. For updated information, create a new version with the same prompt to pull the latest web data.

Understanding when to use research mode versus labs mode helps choose the right tool for different tasks.

Research mode vs labs mode in Perplexity

Research and labs serve different purposes within Perplexity’s workflow. Research mode excels at generating comprehensive written analysis, while labs mode creates interactive applications and visualizations.

When to use Perplexity research mode:

• Literature reviews with citations
• Market analysis and competitive research
• Comprehensive written reports with source verification
• Quick turnaround needs (under 5 minutes)

When to use Perplexity labs mode:

• Interactive dashboards and visualizations
• Web applications with filters and search
• Projects combining multiple data sources
• Tools requiring dynamic content display

Combining both Perplexity modes: Use Research for analytical depth and citations, then labs to transform data into interactive dashboards. The market research project demonstrates this two-phase approach.

Both modes require Perplexity Pro ($20/month) with unlimited Research queries and 50 labs queries monthly.

Conclusion

Perplexity AI combines real-time web search with project creation through two modes:

• Research mode generates written reports with citations in 3-5 minutes for academic reviews, market analysis, and competitive research
• Labs builds interactive dashboards and web applications in 10-30 minutes without coding
• Combining both modes produces the best results: Research provides analytical depth, labs transforms data into interactive tools
• The platform consolidates research, visualization, and development into single workflows

The three projects demonstrate practical applications across literature reviews, market research dashboards, and news monitoring, eliminating the need for separate tools.

To learn more about Perplexity’s research capabilities, explore Codecademy’s free Intro to AI Research with Perplexity course. The course covers search modes, citation systems, Focus modes for targeted research, and collaborative features in Spaces.

Frequently asked questions

1. What are Perplexity Labs?

Labs is a Perplexity Pro feature that creates interactive web applications, dashboards, and visualizations from text prompts, taking 10-30 minutes per project.

2. Is Perplexity the same as ChatGPT?

No. Perplexity searches real-time web data and provides cited sources, while ChatGPT generates responses from training data without live web access.

3. Can you create projects in Perplexity?

Yes. research mode creates written reports with citations, and labs mode builds interactive dashboards, web apps, and visualizations.

4. Which is better, Google or Perplexity?

Perplexity provides structured, analyzed answers with citations. Google returns link lists requiring manual research. Different use cases favor different tools.

5. Can Perplexity Labs make PPT?

No. Labs generates web-based dashboards and applications using HTML, CSS, and JavaScript, not PowerPoint presentations.