Every process is classified by AI relationship (how AI and humans work together) and delivery method (how you implement it). Scored across 6 dimensions.
AI Classification — how AI and humans work together
A
Agentic AI
AI executes multi-step workflows autonomously. Humans monitor outcomes and intervene on exceptions. Includes both ML-driven and rule-based automation.
Human monitors & intervenes on exceptions
B
Augmented AI
AI drafts, recommends, or analyzes. A human reviews and approves every output before action. Right for judgment-heavy, compliance-sensitive, or relationship-driven work.
Human reviews & approves every AI output
E
Human-Led + AI Assist
Human owns the decision — it requires judgment, creativity, or relationship context that AI cannot replicate. AI prepares inputs, drafts outputs, summarizes background, and automates surrounding workflow.
Human owns the decision; AI supports with prep, drafts & automation
Delivery Method — how you implement it
Activate
Capability exists in a platform you already own (Salesforce, SAP, ServiceNow, etc). Configure and turn on — fastest path to ROI.
Build
Custom AI solution needed. Build with LLMs, ML models, or AI agents tailored to your process and data.
Build + Automate
Fully deterministic, high volume, few exceptions, commodity work. Build rule-based automation — no ML needed. Reserved for truly predictable processes.
Example process card
A
Invoice Validation & Billing Compliance Check
Agentic AIActivate · TeamConnect
AI class describes how AI and humans collaborate. Delivery tag describes how you implement it — this process activates AI already in TeamConnect.
Each process is scored 1-4 across six dimensions. Click any dimension to see what each score means. DL=1 forces Human-Led classification regardless of other scores.
DSData Structure
How structured and machine-readable is the input data?
Hi! I'm the HypersightAI Assistant. Ask me about any process, scenario, or AI transformation topic. I can see what you're viewing and provide context-specific guidance.