🤖 GCN AI Worker API

Production-ready AI API for startup assistance, lead scoring, and content generation

Overview

Base URL: https://ai.gcnlabs.com

The GCN AI Worker API provides intelligent services for startups and investors, including:

Guardrails: All endpoints enforce strict input/output limits to control costs and ensure quality. Summaries and bios are limited to 600 characters. The system detects and prevents oversized content generation requests.
🧪 Interactive Testing 📄 OpenAPI Spec (JSON) 📄 OpenAPI Spec (YAML)

Authentication

No authentication is required for public endpoints. API keys are configured server-side. The API automatically routes requests to available providers (OpenAI/Anthropic) based on configuration.

Endpoints

GET /health

Purpose: Check service health and availability

Response:

{
  "ok": true,
  "service": "ai.gcnlabs.com",
  "timestamp": 1234567890
}

GET /models

Purpose: List all available AI models and their capabilities

Response:

{
  "ok": true,
  "models": [
    {
      "key": "gpt-4o-mini",
      "provider": "openai",
      "defaultFor": ["summary", "bio"],
      "costTier": "low"
    },
    {
      "key": "gpt-4o",
      "provider": "openai",
      "defaultFor": ["score"],
      "costTier": "medium"
    },
    {
      "key": "claude-3.5-sonnet",
      "provider": "anthropic",
      "defaultFor": [],
      "costTier": "high"
    }
  ]
}

POST /score/lead

Purpose: Score a lead across 4 dimensions (pitch, booth, ticket, premium)

Default Model: gpt-4o

Request Body:

{
  "lead": {
    "name": "Alice Johnson",
    "company": "TechStartup Inc",
    "role": "Founder & CEO",
    "raiseAmount": 500000,
    "stage": "Seed",
    "industry": "Fintech",
    "linkedin": "https://linkedin.com/in/alice",
    "website": "https://techstartup.com"
  },
  "model": "gpt-4o" // optional
}

Response:

{
  "ok": true,
  "model": "gpt-4o",
  "scores": {
    "pitchScore": 7.5,
    "boothScore": 4,
    "ticketScore": 8,
    "premiumScore": 6,
    "reasoning": {
      "pitch": "Early-stage fintech with clear fundraising need and strong founder.",
      "booth": "Limited B2B service offering reduces booth appeal.",
      "ticket": "Strong professional profile makes networking valuable.",
      "premium": "Moderate credentials; some premium value but not top tier."
    }
  }
}

Example curl:

curl -X POST https://ai.gcnlabs.com/score/lead \
  -H "Content-Type: application/json" \
  -d '{
    "lead": {
      "name": "Alice Johnson",
      "company": "TechStartup Inc",
      "raiseAmount": 500000
    }
  }'

POST /score/investor

Purpose: Score an investor across 6 dimensions with overall score

Default Model: gpt-4o

Scoring Categories:

  • capitalStrength - Check size, fund size, liquidity
  • dealVelocity - Investment frequency, recent deals, portfolio growth
  • sectorAlignment - Industry overlap with ecosystem trends
  • experienceTrackRecord - Years investing, portfolio quality, operator experience
  • networkStrength - LinkedIn connections, warm intros, associations
  • reputationSignal - Credibility, professionalism, trust indicators

Request Body:

{
  "investor": {
    "name": "John Smith",
    "type": "Angel Investor",
    "checkSize": 50000,
    "preferredStages": ["Pre-seed", "Seed"],
    "industries": ["Fintech", "SaaS"],
    "location": "San Francisco",
    "linkedIn": "https://linkedin.com/in/johnsmith",
    "dealsLastYear": 8,
    "portfolioCompanies": 15
  },
  "model": "gpt-4o" // optional
}

Response:

{
  "ok": true,
  "model": "gpt-4o",
  "scores": {
    "overallInvestorScore": 7.8,
    "scores": {
      "capitalStrength": 8,
      "dealVelocity": 7,
      "sectorAlignment": 9,
      "experienceTrackRecord": 7,
      "networkStrength": 8,
      "reputationSignal": 7
    },
    "reasoning": {
      "capitalStrength": "Strong check size suitable for seed rounds",
      "dealVelocity": "Active investor with 8 deals last year",
      "sectorAlignment": "Excellent fit with fintech and SaaS focus",
      "experienceTrackRecord": "Solid portfolio of 15 companies",
      "networkStrength": "Strong LinkedIn presence and warm intro potential",
      "reputationSignal": "Well-regarded in SF ecosystem",
      "overall": "Highly qualified investor with strong capital and relevant sector focus"
    }
  }
}

Example curl:

curl -X POST https://ai.gcnlabs.com/score/investor \
  -H "Content-Type: application/json" \
  -d '{
    "investor": {
      "name": "John Smith",
      "type": "Angel Investor",
      "checkSize": 50000
    }
  }'

POST /score/startup

Purpose: Generate GDF Investability Score Matrix across 8 dimensions

Default Model: gpt-4o

Scoring Categories:

  • team - Experience, founder-market fit, execution track record
  • traction - Revenue, growth, engagement, pilots, customer validations
  • market - TAM, SAM, trend momentum, adoption readiness
  • competitiveAdvantage - Moat, differentiation, IP, barriers to entry
  • businessModel - Clarity, feasibility, pricing strength, scalability
  • pitchQuality - Narrative strength, clarity, pitch deck completeness
  • profileCompleteness - Thoroughness of GDF profile (fields filled, materials uploaded)
  • riskLevel - Team risk, execution risk, regulatory risk (lower risk = higher score)

Request Body:

{
  "startup": {
    "name": "TechVenture AI",
    "industry": "AI/ML",
    "stage": "Seed",
    "revenue": 100000,
    "growth": "15% MoM",
    "teamSize": 5,
    "founderExperience": "Former Google engineers",
    "marketSize": "10B TAM",
    "competitiveAdvantage": "Proprietary ML models",
    "customers": 50,
    "pitchDeckComplete": true
  },
  "model": "gpt-4o" // optional
}

Response:

{
  "ok": true,
  "model": "gpt-4o",
  "scores": {
    "overallInvestabilityScore": 8.2,
    "scores": {
      "team": 9,
      "traction": 7,
      "market": 9,
      "competitiveAdvantage": 8,
      "businessModel": 7,
      "pitchQuality": 8,
      "profileCompleteness": 8,
      "riskLevel": 7
    },
    "reasoning": {
      "team": "Exceptional team with Google engineering experience",
      "traction": "Strong revenue growth at 15% MoM with 50 customers",
      "market": "Large 10B TAM in hot AI/ML space",
      "competitiveAdvantage": "Proprietary ML models provide defensibility",
      "businessModel": "Clear B2B SaaS model with recurring revenue",
      "pitchQuality": "Complete pitch deck with compelling narrative",
      "profileCompleteness": "Comprehensive GDF profile with key metrics",
      "riskLevel": "Moderate risk due to competitive landscape",
      "overall": "Highly investable startup with exceptional team and strong traction in large market"
    }
  }
}

Example curl:

curl -X POST https://ai.gcnlabs.com/score/startup \
  -H "Content-Type: application/json" \
  -d '{
    "startup": {
      "name": "TechVenture AI",
      "stage": "Seed",
      "revenue": 100000
    }
  }'

POST /ai/summary

Purpose: Generate a concise startup summary (max 600 characters)

Default Model: gpt-4o-mini

Request Body:

{
  "startupName": "Acme AI",
  "description": "We build AI-powered customer service automation tools for small businesses...",
  "industry": "AI/ML",
  "stage": "Seed",
  "model": "gpt-4o-mini" // optional
}

Response:

{
  "ok": true,
  "model": "gpt-4o-mini",
  "summary": "Acme AI is a platform that provides AI-powered customer service automation for small businesses. The company helps SMBs reduce response times and improve customer satisfaction through intelligent chatbots and workflow automation. Acme AI is at the Seed stage and serves the AI/ML industry."
}

Example curl:

curl -X POST https://ai.gcnlabs.com/ai/summary \
  -H "Content-Type: application/json" \
  -d '{
    "startupName": "Acme AI",
    "description": "AI-powered customer service automation for SMBs"
  }'

POST /ai/bio

Purpose: Generate a founder bio in third person (max 600 characters)

Default Model: gpt-4o-mini

Request Body:

{
  "name": "Jane Doe",
  "role": "Founder & CEO",
  "background": "Former Google engineer with 10 years in ML, Stanford CS grad",
  "company": "Acme AI",
  "tone": "professional", // optional: "professional", "casual", "investor-facing"
  "model": "gpt-4o-mini" // optional
}

Response:

{
  "ok": true,
  "model": "gpt-4o-mini",
  "bio": "Jane Doe is the Founder & CEO of Acme AI, bringing over 10 years of machine learning expertise from her time at Google. A Stanford Computer Science graduate, she leads the company's vision to democratize AI-powered customer service for small businesses."
}

Example curl:

curl -X POST https://ai.gcnlabs.com/ai/bio \
  -H "Content-Type: application/json" \
  -d '{
    "name": "Jane Doe",
    "role": "Founder & CEO",
    "company": "Acme AI"
  }'

POST /ai/image

Purpose: Generate images using OpenAI DALL-E models

Default Model: dall-e-3

Request Body:

{
  "prompt": "A professional logo for a tech startup",
  "model": "dall-e-3", // optional: "dall-e-2", "dall-e-3"
  "size": "1024x1024", // optional: "1024x1024", "1792x1024", "1024x1792"
  "quality": "standard" // optional: "standard", "hd"
}

Response:

{
  "ok": true,
  "model": "dall-e-3",
  "size": "1024x1024",
  "quality": "standard",
  "prompt": "A professional logo for a tech startup",
  "images": [
    {
      "url": "https://...",
      "revised_prompt": "A modern, minimalist logo design..."
    }
  ]
}

Example curl:

curl -X POST https://ai.gcnlabs.com/ai/image \
  -H "Content-Type: application/json" \
  -d '{
    "prompt": "A futuristic office building",
    "size": "1792x1024",
    "quality": "hd"
  }'

Note: Requires OPENAI_IMAGE_API_KEY environment variable. DALL-E 3 only supports generating 1 image at a time.

POST /ai/generate

Purpose: Generic generation endpoint with flexible task types and model selection

Request Body:

{
  "task": "summary", // "summary", "bio", or "custom"
  "input": {
    "description": "A fintech startup that helps people save money..."
  },
  "model": "gpt-4o-mini" // optional
}

Response:

{
  "ok": true,
  "task": "summary",
  "model": "gpt-4o-mini",
  "output": "Generated content based on the task and input..."
}

Example curl:

curl -X POST https://ai.gcnlabs.com/ai/generate \
  -H "Content-Type: application/json" \
  -d '{
    "task": "custom",
    "input": {"prompt": "Explain the benefits of angel investing"},
    "model": "claude-3.5-sonnet"
  }'

Model Selection

You can override the default model for any endpoint by including a model parameter in your request:

If a requested model's API key is not configured, you'll receive a clear error message suggesting an alternative.

Error Handling

All endpoints return consistent error responses:

{
  "ok": false,
  "error": "Description of the error"
}

Common HTTP Status Codes:

CORS Support

All endpoints support CORS with the following headers:

OPTIONS preflight requests are handled automatically.

Rate Limits & Guardrails

Input Limits:
  • Summary/bio descriptions: 600 characters max
  • Lead data: 8,000 characters max (JSON stringified)
  • Custom generation prompts: 1,000 characters max
Output Limits:
  • Summaries and bios: 600 characters max
  • Custom generation: 1,000 characters max
  • All outputs are automatically truncated if they exceed limits
Dangerous Request Detection:

The system automatically detects requests for oversized content (e.g., "write a full business plan") and converts them into safe, bounded tasks.

Machine-Readable Specs

For automated integration and tool discovery:

These specs can be used with:

Support & Contact

For questions or support, contact GCN Labs: