supplygraph-ai

Enterprise Supply Graph Agent

Overview

The Enterprise Supply Graph Agent generates a multi-tier, company-centric global supply chain graph for any enterprise in the world.

It automatically maps relationships across:

This agent transforms fragmented supply networks into a living, explorable, and actionable graph — enabling organizations to see, understand, and manage their real dependencies for the first time.

Pain

Most companies operate with severely limited visibility into their true supply networks.

Without deep-tier transparency:

Traditional tools only reveal Tier-1 suppliers at best — leaving 80–90% of the real risk surface invisible.

Breakthrough

The Enterprise Supply Graph Agent solves this by:

For the user, the experience is simple:

  1. Enter a company name
  2. Confirm the correct entity
  3. Instantly visualize an interactive, multi-tier supply network

What was once invisible becomes navigable, measurable, and actionable.

Why SupplyGraph AI

Powered by:

SupplyGraph AI delivers the first truly intelligent, real-time, multi-tier enterprise supply graph in the world — turning hidden dependency structures into strategic advantage.

Try the Enterprise Supply Graph Agent (Live Chatbot)

Before integrating this agent via API, you can experience it directly through our interactive visualization chatbot.

This live demo allows you to:

Launch the Enterprise Supply Graph Chatbot
https://supplygraph.ai/zk_chat_os/agentic/dialog.html?name=sg_visualization

To use the chatbot, you’ll first need to:

This chatbot is powered by the same Enterprise Supply Graph Agent and A2A endpoints described in this documentation.
Credits used in the chatbot are deducted in the same way as API / A2A usage.

Everything you experience here can be fully embedded into your own system through A2A integration.

Sandbox Key Support (for Development)

This agent supports Sandbox API Keys, allowing testing without consuming credits.

When using a Sandbox Key:

This is useful for:

⚠️ Sandbox datasets are static and do not represent real-time updates.
Use Production Keys for live graph computation.

For details, see
Getting Started Guide → API Keys.

Input Requirements

agent_id: sg_visualization · MCP tool: sg_visualization · Pricing: 264590 credits / run

Input differs by integration surface:

A2A & Agent API

Field Required Description
text Yes Natural-language company name (e.g. "Tesla, Inc.")

Example: "Tesla, Inc."

The agent resolves the company name and may ask you to confirm the matched entity before graph computation starts (see Multi-turn Example (A2A)).

MCP

MCP does not use multi-turn company confirmation on sg_visualization. Resolve the company ID first, then pass pid directly.

Step MCP tool Input Description
1 search_company_candidates text Company name with optional country/region (e.g. "Tesla United States")
2 sg_visualization pid Internal company ID from step 1 (candidates[].pid)

Step 1 example: {"text": "Tesla United States"}
Step 2 example: {"pid": "a77828f060c866441f2403384b271e63"}

search_company_candidates pricing: 1 credit / run. See MCP search tool output below.

Full machine-readable schemas → GET /api/v1/agents/sg_visualization/manifest (see Agent API §3).

Output Format

Primary output lives in data.content (Agent API) or task artifacts (A2A / MCP). The payload is a oneOf:

State content type Description
In progress, failed, cancelled, or waiting for user input string Text or Markdown — validation prompts, confirmation requests, or error messages
Task completed successfully object Structured supply chain graph (see below)

Structured success object:

{
  "type": "sg_visualization_results",
  "data": {
    "pid": "a77828f060c866441f2403384b271e63",
    "company_name": "Tesla, Inc.",
    "max_depth": 3,
    "dependencies": [
      { "from": "0->-1", "to": "1->1384787", "dep_depth": 1 }
    ],
    "node_info": {
      "1->1384787": {
        "name": "HBM Memory",
        "index_values": [
          { "index_name": "Number of enterprises", "value": 13893.0 }
        ],
        "company_distribution": [
          { "country": "China", "count": 1213 }
        ]
      }
    }
  }
}
Field Description
type Always "sg_visualization_results" on success
data.pid Unique ID of the company-centered supply chain graph
data.company_name Name of the target company
data.max_depth Maximum dependency depth (longest path in the hierarchy)
data.dependencies[] Graph edges — from / to use "level->nodeID" format; dep_depth is edge depth
data.node_info Map of node ID → { name, index_values[], company_distribution[] }

Estimated task duration: 30 minutes – 7 hours (Production, live graph build). Sandbox returns instantly.

After entity confirmation (A2A) or pid submission (MCP), the task enters a long-running executing state — poll status until TASK_COMPLETED, then call results.

Sample Response (Sandbox)

Sandbox returns the Tesla, Inc. fixture with the same structure as Production. Graph data below is abbreviated.

Success (TASK_COMPLETED) — Agent API results:

{
  "success": true,
  "code": "TASK_COMPLETED",
  "message": "Task completed successfully.",
  "data": {
    "task_id": "<task-id>",
    "agent": "sg_visualization",
    "stage": "completed",
    "progress": 100,
    "content": {
      "type": "sg_visualization_results",
      "data": {
        "pid": "a77828f060c866441f2403384b271e63",
        "company_name": "Tesla, Inc.",
        "max_depth": 3,
        "dependencies": [
          { "from": "0->-1", "to": "1->1384787", "dep_depth": 1 },
          { "from": "0->-1", "to": "1->1683397", "dep_depth": 1 },
          { "from": "0->-1", "to": "1->1657001", "dep_depth": 1 }
        ],
        "node_info": {
          "1->1384787": {
            "name": "HBM Memory",
            "index_values": [
              { "index_name": "Number of enterprises", "value": 13893.0 },
              { "index_name": "Number of enterprise software copyrights", "value": 2728.0 }
            ],
            "company_distribution": [
              { "country": "China", "count": 1213 },
              { "country": "Japan", "count": 14 }
            ]
          }
        }
      }
    }
  },
  "metadata": { "credits_used": 0 },
  "errors": null
}

Multi-turn Example (A2A)

A2A and Agent API only. MCP uses the two-tool workflow (search_company_candidates → sg_visualization) instead of multi-turn confirmation.

When the agent cannot unambiguously match a company from text, it returns WAITING_USER with a confirmation prompt in content (string). Continue with the same task_id.

Turn 1 — submit company name:

curl -X POST "https://agent.supplygraph.ai/api/v1/agents/sg_visualization/run" \
  -H "Authorization: Bearer <YOUR_API_KEY>" \
  -H "Content-Type: application/json" \
  -d '{"text": "Tesla, Inc.", "stream": false}'

Poll status until code is WAITING_USER. Example results content:

{
  "success": true,
  "code": "WAITING_USER",
  "data": {
    "task_id": "<task-id>",
    "content": "Here are the companies we've identified.\nCompany Name: Tesla, Inc.\nCountry: United States\nPlease reply [Yes] or [No] to confirm."
  }
}

Turn 2 — confirm entity (same task_id):

curl -X POST "https://agent.supplygraph.ai/api/v1/agents/sg_visualization/run" \
  -H "Authorization: Bearer <YOUR_API_KEY>" \
  -H "Content-Type: application/json" \
  -d '{"text": "Yes", "task_id": "<task-id>", "stream": false}'

The agent accepts the match, deducts credits, and queues graph computation. Poll status until TASK_COMPLETED (may take 30 min – 7 h in Production), then retrieve structured sg_visualization_results via results.

See Agent API §8 for the general multi-turn pattern.

MCP Company Resolution (search_company_candidates)

Use this MCP tool before sg_visualization to resolve a company name into a pid. This tool is single-turn (no confirmation step).

Field Required Description
text Yes Company name with optional country/region

Success — structured output:

{
  "type": "company_candidate_search",
  "data": {
    "candidates": [
      {
        "pid": "a77828f060c866441f2403384b271e63",
        "company_name": "Tesla, Inc.",
        "country_name": "United States"
      }
    ]
  }
}

Select the appropriate pid, then call sg_visualization:

# 1. Resolve company ID
search_result = await session.experimental.call_tool_as_task(
    "search_company_candidates",
    {"text": "Tesla United States"},
)
# ... poll until complete, parse candidates[0]["pid"]

# 2. Build supply graph (no multi-turn)
graph_result = await session.experimental.call_tool_as_task(
    "sg_visualization",
    {"pid": "a77828f060c866441f2403384b271e63"},
)

Estimated duration for search_company_candidates: 5–8 seconds.

Integration

Method ID / Tool Documentation
A2A sg_visualization a2a.md
MCP search_company_candidates + sg_visualization mcp.md
Agent API sg_visualization agent-api.md

Call pattern is identical across agents — only agent_id / tool name and input differ.

Quick examples: A2A / MCP · Agent API

Errors

Common codes → Agent API §10.

Situation Code Agent behavior
Ambiguous or unmatched company (A2A / Agent API) WAITING_USER Prompts for entity confirmation; resume with same task_id
Invalid or unknown pid (MCP) TASK_FAILED Returns guidance in content (string) — re-run search_company_candidates
Insufficient credits INSUFFICIENT_CREDITS Top up via Console
Graph still building TASK_RUNNING Continue polling status; Production builds may take up to 7 hours

Maintainer: info@supplygraph.ai
License: Proprietary / Internal
© 2025–2026 SupplyGraph AI. All rights reserved.

This site is open source. Improve this page.