Inside Instirio, the engine is Halia v1.0

Eight modules. One engine. One job: find the money.

Eight modules, one detection engine, grounded in three industry standards — SCOR, IEEE XES 1849, ISO/IEC 25012.

Standards-grounded math. No black-box proprietary scoring.

Aligned with SCOR · IEEE XES 1849 · ISO/IEC 25012
Onboarding journey

From empty to insightful in five minutes.

Day 1: connect a platform, see what we’re going to find. Day 30: a populated dashboard with the issues and dollar amounts that matter.

Day 1 · empty

Connect Shopify to begin.

C
Instirio dashboard
No platforms connected

Connect your first platform

Pick from Shopify, WooCommerce, ShipStation, Stripe, Amazon, or ShipBob. Setup takes 60 seconds.

Issues to fix
Connect to populate
Money at risk
Connect to populate
Order speed
Connect to populate
Money saved
Connect to populate
  • 1
    Pick a platform. OAuth or API key, 60 seconds.
  • 2
    Configure costs. Pick your region; we pre-fill regional benchmarks.
  • 3
    Wait 5 minutes. Historical events sync; first findings appear.
Day 30 · populated

Findings ranked by dollar impact.

C
Instirio dashboard
Live · 3 platforms
Top finding · Critical
Return rate jumped to 12.4%
$5,541
Issues to fix
13
+2 vs last wk
Money at risk
$28,940
+$3.2K wk
Order speed
42.8h
−6.2h base
Money saved
$14,380
2 fixed
Live 11 of 12 metrics stable · 1 drift detected
  • 37 detectors running. Findings ranked by dollar impact.
  • Drift monitor live. Regressions caught within an hour.
  • SCOR metrics computed. Perfect Order Rate, Cost to Serve, Cycle Time on every dashboard load.
Module 1

How does Instirio visualise your order flow?

Every order’s flow through every state, auto-laid-out. Bottlenecks glow red, returns branch below. Click any node for transitions and timing.

  • 16-event canonical taxonomy normalized across Shopify, WooCommerce, ShipStation, Stripe, Amazon, ShipBob.
  • MAX(incoming, outgoing) node counts, prevents the inflated counts you see in other tools.
  • Repeat-purchase filter, Delivered→Order Created loops excluded so the map reflects the actual order lifecycle.
  • Bidirectional merge, 0-delay pairs collapsed into the forward direction for clarity.
MAIN FLOW Order created 2,140 Payment 2,108 Inventory 2,096 Packing 2,090 27.6h Shipped 2,070 Delivered 1,980 EXCEPTION FLOW Return requested 164 12.4% rate Return received 152 Refund processed 148 DETECTOR Return leakage $5,541/mo Bottleneck +27.6h LEGEND Bottleneck Exception Detector finding Drift / live signal Sample: 2,140 orders / 30d · auto-detected from order-event log

Auto-laid-out process map with bottleneck highlighting and exception-flow separation.

Module 2 · powered by Halia

37 detectors. Eight categories. One engine.

Halia runs 37 always-on detectors orchestrating statistics, ML, and graph reasoning. Every finding ships with a dollar amount, severity score, and confidence, so you fix what matters.

🛡️
9 detectors

Fraud & revenue leakage

Card-testing, serial returner, coupon abuse, refund fraud, chargeback rate, duplicate orders.

📦
4 detectors

Returns & product quality

Defective-batch clustering, reason-only patterns, recall surges, SKU velocity spikes.

🚚
8 detectors

Fulfillment, carrier & SLA

Stuck orders, SLA breach, vendor SLA, carrier-zone drift, partial shipments, BOPIS, viral spikes.

💰
5 detectors

Margin, pricing & cost creep

COGS drift, order-level margin, AOV drift, processor-fee drift, FX timing loss.

🔁
2 detectors

Subscriptions & retention

Subscription churn signals, gift-subscription cancellation early warning.

📋
2 detectors

Inventory & demand

Oversell at checkout (ERP sync lag), extended out-of-stock.

🛠️
7 detectors

Process & operational efficiency

Cancellation leaks, payment friction, rework loops, slow refunds, statistical outliers, trend drift, new variants.

👥
4 detectors

Team & time analytics

Actor performance, time-of-day patterns, channel imbalance, automation opportunities.

Always-on for every plan: Fraud, Returns & SLA detectors. Business+ unlocks Margin, Subscriptions, Inventory, Process and Team tiers. See plan details →

Enterprise-grade rigor: multi-factor severity, repeat-purchase filtering, dynamic impact capping, overlap deduplication, confidence calibration, and detection debouncing, built in by default.
Module 3

SCOR metrics. The standard your industry uses.

Fortune 500 supply-chain leaders use the SCOR Digital Standard from ASCM. Instirio brings it to SMB e-commerce, calibrated for your scale.

  • RL
    Perfect Order Rate (RL.1.1), 4-factor multiplication: on-time × in-full × damage-free × doc-accuracy.
  • RS
    Cycle Time (RS.1.1), Source (order→payment) + Make (payment→shipped) + Deliver (shipped→delivered).
  • CO
    Cost to Serve (CO.1.1), 5-category breakdown: Plan + Source + Make + Deliver + Return.
Perfect Order Rate
94.2%
on-time 97.8% × in-full 99.1% × dmg-free 98.4% × doc 99.0%
Cycle Time
42.8h
Source 4.2h Make 18.4h Deliver 20.2h
Cost to Serve · per order
$129.76
Plan $1.50 Source $88.35 Make $16.00 Deliver $20.19 Return $3.72
Drift Avg lead time +18.2% vs 30-day baseline (42.8h → 50.6h)
Regression Return rate 12.4%, fixed in March, baseline 8.1%
Stable Perfect Order Rate 94.2%, within ±1.2% baseline tolerance
Module 4

Drift detection, catches regressions other tools miss.

Most analytics tools show what’s happening now. Instirio baselines every metric, then flags the moment things drift, including fixes that quietly come undone.

  • Per-team baselines that learn your normal pace. Updated continuously. No manual configuration.
  • Regression flagging. Drift events tagged as regressions when they cross the original detector threshold again.
  • Transition baselines. Step-by-step delay drift, not just aggregate metrics.
Module 5

How does Instirio track SLA compliance?

Define rules, track compliance, attach a cost-per-breach. Get notified before Amazon’s 4% SLA threshold suspends your listings.

  • Rules CRUD, create/edit/delete with templates and auto-generation from your historical data.
  • Compliance bars, daily compliance % with 14-day trend.
  • Cost-per-breach. Every breach tagged with dollar impact.
  • Cost configuration, region selector for 11 regions with auto-analyze from logs.
Ship within 48h of order Active
96.4% compliance · 14d 78 breaches · est. cost $1,170
Refund within 5 business days Drifting
81.2% compliance · 14d 28 breaches · est. cost $840
Amazon ship-confirm 1 business day Breach risk
72.8% compliance · 14d Listing suspension threshold: 96%
Order Created
2,140 events · system actor
Automated
Inventory Allocated
2,096 events · WMS
Automated
Quality Check
1,840 events · 8 human actors · irregular timing
Manual
Label Printed
2,070 events · ShipStation batch
Automated
Return Received
152 events · 3 human actors
Manual
Module 6

How does Instirio find manual work to automate?

Multi-factor classification separates system, carrier, and human actors. Temporal patterns reveal which steps are truly automated vs. masquerading as automated.

  • Actor type detection, system, carrier, human, based on event metadata and timing.
  • Batch detection, bursts of identical events from one actor flagged as batch.
  • Drill-down, per-activity automation assessment, hourly heatmap, sample orders.
  • Manual steps banner, surfaces the candidates for automation, ranked by event count.
Module 7

What does an order cost where you operate?

Six cost elements, shipping, packaging, labor, returns, processing, COGS%, benchmarked across 11 regions. Auto-analyzed from your event logs, with volume-weighted multi-region blending.

  • 11 regions, North America, Western Europe, UK, Eastern Europe, South Asia, Southeast Asia, East Asia, Middle East, Latin America, Africa, Australia/NZ.
  • 6 cost elements per region with low/mid/high ranges from public benchmarks (APQC, Opensend, GoBolt, Speed Commerce, TrueProfit).
  • Auto-analyze, derives realistic costs from your shipping events; you confirm or override.
  • Per-connection region tagging so multi-region brands get correct math.
North America benchmark · shipping
$8.40 Low $11.20 Mid $15.80 High
Your store: $20.04/order 2× regional benchmark
$3,200/mo overcharge detected vs. North America regional benchmark. Compare carriers →
Module 8

How does Instirio validate its own data?

Instirio validates its own data. P0 checks run after every sync, P1 hourly, if a number’s wrong, you’ll know before your CFO does.

P0. Every sync

8 checks
  • Events ↔ summaries consistency
  • NaN impact detection
  • SCOR formula validation
  • Lead time accuracy
  • Connection completeness
  • Detector concurrency limits
  • Detection debounce respected
  • Event ordering / timestamp sanity

P1, hourly / on-demand

8 checks
  • Delivered = completed alignment
  • Cost configuration completeness
  • Lead time precision (P95)
  • Drift baseline freshness
  • Detection coverage
  • Cache hit ratio
  • Worker health
  • Backup verification
SCOR
SCOR Digital StandardASCM · POR / CTS / Cycle Time
IEEE
IEEE XES 1849Process-mining log standard
ISO
ISO/IEC 25012Data quality model

Run all eight modules on your data.

Free for 500 orders/mo. Setup in 5 minutes. No credit card.

Start free