Changelog

A record of changes as the system improves — features, enhancements, and fixes, applied continuously.


2.0

HI Engine 2.0 — Execution Layer

You can now assign implementation tasks directly to HI Agent.

Previously, HI Engine generated insights, impact estimates, and implementation tasks. The final implementation step still required manual development work.

With the new Execution Layer, HI Agent can now:

  • Execute approved implementation tasks

  • Generate Shopify theme modifications

  • Create a new draft theme version

  • Prepare changes for review before publishing

  • Provide implementation logs and execution status



Current capabilities

  • Shopify theme modifications

  • CRO implementation tasks

  • Theme section updates

  • UI/UX improvements

  • PDP enhancements

  • Conversion optimization recommendations

Human review

Every execution still includes a human review step before deployment to ensure implementation quality and prevent unwanted changes.

Why it matters

The biggest bottleneck is often not identifying opportunities — it's getting them implemented.

The Execution Layer closes the gap between insight and implementation.

Available now in HI Engine Pro. 🚀

v1.0.2

CRO Dashboard & Root-Cause Investigation

A unified analytics view with AI-powered diagnosis for every metric change.


New

  • CRO Dashboard: Shopify and Google Analytics metrics — conversion rate, revenue, sessions, AOV, funnel drop-offs — side by side in a single view.

  • Root-Cause Investigator: An AI agent automatically diagnoses why a metric moved, using live data from GA4, ShopifyQL, and Clarity session recordings.

  • Super Insight Agent: Combines GA4, ShopifyQL, and behavioral data in a single reasoning pass for faster, more accurate insights.


Improvements

  • Every CRO idea now includes a hypothesis, supporting evidence, and estimated revenue impact

  • Funnel analysis pinpoints exactly where customers drop off from product view to purchase

  • Automatic discrepancy detection when Shopify and GA4 numbers diverge

V1.1.0

Weekly Queue with TRICE Scoring

AI-prioritized weekly batches with deep investigation and admin review.

New

  • TRICE Scoring: Every candidate task is scored on Trust, Revenue-impact, Implementation-cost, Confidence, and Evidence — only the highest-value work surfaces for the week.

  • Deep Investigation: Before scoring, an AI agent pulls live GA4 and Shopify data for each task to ground its recommendation in real numbers.

  • Per-Store Batches: Each connected store gets its own tailored weekly queue.

  • Visibility Toggle: Preview exactly what the customer will see before publishing a batch.

Improvements

  • Title-similarity dedup prevents near-duplicate tasks from appearing across weeks

  • Batch approve, edit inline, or regenerate the entire week in one click

  • Specs are decoupled from implementation-biased titles for cleaner briefs

v1.2.0

Kanban Task Board

A visual board for managing your weekly growth tasks from start to finish.

New

  • Kanban Task Board: Your weekly growth tasks now live on a drag-and-drop board with columns for each lifecycle stage — To Do, In Progress, In Review, and Done.

  • Auto-Clean Board: Dismissed tasks are automatically hidden so the board only shows active work.

  • Task Cards at a Glance: Each card surfaces priority, estimated impact, and assignee without clicking in.

Improvements

  • Approving a task in the weekly queue now syncs it to the board instantly

  • Email invite templates updated to match the orange design system accent