Gamma Core Workflow A: AI Presentation Generation

Overview

Implement the core workflow for generating presentations using Gamma's AI capabilities from various input sources.

Prerequisites

• Completed

  gamma-sdk-patterns

  setup
• Understanding of async patterns
• Content ready for presentation

Instructions

Step 1: Prompt-Based Generation

import { GammaClient } from '@gamma/sdk';

const gamma = new GammaClient({ apiKey: process.env.GAMMA_API_KEY });

async function generateFromPrompt(topic: string, slides: number = 10) {
  const presentation = await gamma.presentations.generate({
    prompt: topic,
    slideCount: slides,
    style: 'professional',
    includeImages: true,
    includeSpeakerNotes: true,
  });

  return presentation;
}

// Usage
const deck = await generateFromPrompt('Introduction to Machine Learning', 8);
console.log('Generated:', deck.url);

Step 2: Document-Based Generation

async function generateFromDocument(filePath: string) {
  const document = await fs.readFile(filePath, 'utf-8');

  const presentation = await gamma.presentations.generate({
    sourceDocument: document,
    sourceType: 'markdown', // or 'pdf', 'docx', 'text'
    extractKeyPoints: true,
    maxSlides: 15,
  });

  return presentation;
}

Step 3: Structured Content Generation

interface SlideOutline {
  title: string;
  points: string[];
  imagePrompt?: string;
}

async function generateFromOutline(outline: SlideOutline[]) {
  const presentation = await gamma.presentations.generate({
    slides: outline.map(slide => ({
      title: slide.title,
      content: slide.points.join('\n'),
      generateImage: slide.imagePrompt,
    })),
    style: 'modern',
  });

  return presentation;
}

Step 4: Batch Generation Pipeline

async function batchGenerate(topics: string[]) {
  const results = await Promise.allSettled(
    topics.map(topic =>
      gamma.presentations.generate({
        prompt: topic,
        slideCount: 5,
      })
    )
  );

  return results.map((r, i) => ({
    topic: topics[i],
    status: r.status,
    url: r.status === 'fulfilled' ? r.value.url : null,
    error: r.status === 'rejected' ? r.reason.message : null,
  }));
}

Output

• AI-generated presentations from prompts
• Document-to-presentation conversion
• Structured content transformation
• Batch processing capability

Error Handling

Error	Cause	Solution
Generation Timeout	Complex prompt	Reduce slide count or simplify
Content Too Long	Document exceeds limit	Split into sections
Rate Limit	Too many requests	Implement queue system
Style Not Found	Invalid style name	Check available styles

Resources

• Gamma AI Generation
• Prompt Best Practices

Next Steps

Proceed to

gamma-core-workflow-b

for presentation editing and export workflows.