Features · Data Modeling & Analysis

Turn device data into AI-executable assets

One JavaScript model connects device data, RPC, triggers, and data assets. AI Agents use domain Skills to implement modeling, automation, and dashboards.

ThinkLink thing-model management with model records, import, export, and version access
AI MODELJS + tklHelperinline RPC / Trigger
Real platform UI · Model management

One data workflow, from intent to presentation

01 · DESCRIBE

Describe the outcome

Provide the device protocol, sample payload, and business goal.

02 · MODEL

Model with AI

Generate a verifiable model with JavaScript, tklHelper, and Skills.

03 · ASSET

Organize data assets

Aggregate device, gateway, and weather data by relation and location.

04 · PRESENT

Let the Agent implement

Invoke the cockpit or cardlist Skill to implement the dashboard.

AI-NATIVE MODELING

All JavaScript. One modeling context.

The AI Agent reads protocols, sample payloads, and business goals, then uses ThinkLink Skills to improve first-pass usability of the generated model.

  • Use the tklHelper domain toolkit without imports
  • Inject extensions through plugins and env
  • Inline RPC and triggers in the shared model context
  • Verify raw data, parsed output, execution logs, and versions
ThinkLink thing-model list with detail and version-history actions
Real platform UI · JavaScript model management
ThinkLink trigger-model management interface
Real platform UI · Trigger models
ThinkLink location management with list and hierarchy-tree views
Real platform UI · Location management
DATA ASSETS

Organize scattered device data into regional assets

Device, gateway, weather, and other external data first enter a shared modeling workflow, then form new park, building, and floor-level data through relations and location hierarchy.

  • Virtual assets aggregate regional data through device relations
  • Recalculate from the latest scope when relations change
  • Give every device hierarchical location context
  • Include gateway, weather, and other MQTT data in asset modeling
AGGREGATERegional temperature and humidity · Total energy · Online rate
AGENT DASHBOARD

Describe the goal. Let the Agent invoke the right Skill.

The Agent reads model fields, aggregated assets, and location relations, then implements a dashboard with the Skill suited to the business goal and continues refining it in natural language.

  • cockpit Skill for operations centers, park overview, and control rooms
  • cardlist Skill for inspections, status lists, and mobile summaries
  • Keep refining metrics, filters, layout, and visual hierarchy
COCKPIT SKILL
ThinkLink dashboard showing live device metrics and a control action
Real platform UI · Dashboard
CARDLIST SKILL

Device status inspection

  • Temperature sensor · Normal
  • 4G DTU · Online
  • Energy node · Needs attention

AGENT FLOW

  1. Describe the business goal
  2. Read model, asset, and location context
  3. Invoke the cockpit / cardlist Skill
  4. Implement and refine the dashboard

Ready to turn device data into executable assets?