Skip to main content

​
Generate Structured Object

Extract or generate structured data from natural language using AI. Perfect for parsing unstructured text, form filling, data extraction, and more.
Best for Extracting Sales Intelligence: Combine with services.scrape.website() to extract structured information from prospect websites. Perfect for questions like “What products does this company sell?” or “How many decision-makers are listed on their about page?”

​
Method

services.ai.generateObject(params);

​
Parameters

​
prompt
string
required
The prompt describing what object to generate or what data to extract
​
schema
any
required
The schema defining the structure of the output. Accepts a Zod schema instance or a plain JSON Schema object
​
model
string
default:"gpt-5-mini"
The AI model to use. Options: - "gpt-5-mini" - Fast and cost-effective (default) - "gemini-2.0-flash" - Google’s fast model - "gemini-2.5-pro" - Google’s most capable model - "claude-sonnet-4-5-20250929" - Anthropic’s latest model

​
Returns

Returns a Promise with:
​
object
object
required
The generated JSON object matching the provided schema

​
Examples

​
Extract Lead Information from Email Signature

// Extract structured lead data from an email signature or bio
const emailSignature = ctx.thisRow.get("Email Signature");

const result = await services.ai.generateObject({
   prompt: `Extract contact info from this email signature: ${emailSignature}`,
   schema: z.object({
      name: z.string(),
      email: z.string(),
      phone: z.string().optional(),
      company: z.string(),
      title: z.string(),
      linkedin: z.string().optional(),
   }),
});

ctx.thisRow.set({
   Name: result.object.name,
   Email: result.object.email,
   Phone: result.object.phone,
   Company: result.object.company,
   Title: result.object.title,
});

​
Extract Company’s Product Offerings

// Scrape and extract what products/services a prospect company offers
const website = ctx.thisRow.get("Website");
const scraped = await services.scrape.website({ url: website });

const result = await services.ai.generateObject({
   prompt: `Extract the company's product/service offerings from this website: ${scraped.markdown}`,
   schema: z.object({
      primary_products: z.array(z.string()),
      target_market: z.string(),
      pricing_model: z.string().optional(),
      key_features: z.array(z.string()),
      use_cases: z.array(z.string()).optional(),
   }),
});

ctx.thisRow.set({
   Products: result.object.primary_products.join(", "),
   "Target Market": result.object.target_market,
   "Pricing Model": result.object.pricing_model,
});

​
Extract Company Intelligence from About Page

Recommended Pattern: This is the preferred approach for enriching prospect data from their website. It’s more accurate and cost-effective than using web search for targeted data extraction.
// Scrape prospect's about page for company intelligence
const website = ctx.thisRow.get("Website");
const scraped = await services.scrape.website({
   url: `${website}/about`,
});

// Extract structured company data for sales context
const result = await services.ai.generateObject({
   prompt: `Extract company information from this about page: ${scraped.markdown}`,
   schema: z.object({
      company_name: z.string(),
      founded_year: z.number().optional(),
      headquarters: z.string().optional(),
      employee_count_estimate: z.string().optional(),
      description: z.string(),
      key_executives: z
         .array(
            z.object({
               name: z.string(),
               title: z.string(),
            })
         )
         .optional(),
      mission_statement: z.string().optional(),
   }),
});

ctx.thisRow.set({
   "Company Name": result.object.company_name,
   Founded: result.object.founded_year,
   "HQ Location": result.object.headquarters,
   "Company Description": result.object.description,
});

​
Identify Decision Makers from Team Page

// Extract decision makers from a prospect's team/about page
const website = ctx.thisRow.get("Website");
const scraped = await services.scrape.website({
   url: `${website}/team`,
});

const result = await services.ai.generateObject({
   prompt: `Extract all executives and decision makers from this team page. 
   Focus on C-level, VPs, and Directors: ${scraped.markdown}`,
   schema: z.object({
      total_decision_makers: z.number(),
      decision_makers: z.array(
         z.object({
            name: z.string(),
            title: z.string(),
            department: z.string().optional(),
            linkedin_url: z.string().optional(),
         })
      ),
   }),
});

ctx.thisRow.set({
   "Decision Maker Count": result.object.total_decision_makers,
   "Key Contacts": result.object.decision_makers.map((dm) => `${dm.name} (${dm.title})`).join("; "),
});