Back to all articles
iCommerceAI AgentsOperations

From eComm to iComm:
What Really Changed

"If‑this‑then‑that" got us started. Now, commerce needs "if‑this‑then‑learn."

Dominic Steil

Dominic Steil

Founder & CEO at StateSet
Sep 3, 20257 min read

For years, automation meant rigid rules, triggers, and integrations. Useful, until growth added new channels, campaigns, and countless edge cases. Then brittle rules started breaking. Teams pivoted from strategy to firefighting: patching flows, maintaining playbooks, and chasing exceptions.

Today, a foundational change is here: Agentic Operations. These are workflows that learn, reason, and adapt. Instead of "set and forget," agents observe outcomes, improve continuously, and escalate to humans only when nuanced judgment is truly required.

The iCommerce Impact

Reduction in manual tasks
Faster resolution times
ROI in first year
Time to implementation

A Look Back: From Brittle Automation to Adaptive Agents

(Bias acknowledged, I built a lot of systems this way.) A state change triggered an event (often a webhook with a JSON payload) that hits an async serverless API to update state in another system. We wrote logic, produced an output, and updated state. Classic commerce ops patterns included:

  • Generating a return label and emailing a PDF to the customer.
  • Processing a refund in Stripe after calculating deductions based on item condition.
  • Looping through non‑returns and generating Stripe invoices.

In short: events → endpoint logic → state update → return. It worked, until it didn't. The GPT wave changed everything.

Code Evolution: From Rules to Reasoning

Traditional Approach

// Old Way: Brittle Rule-Based System
if (return.reason === 'damaged') {
  if (return.value > 100) {
    sendToManager();
  } else {
    autoApprove();
    generateLabel();
  }
} else if (return.reason === 'not_as_described') {
  // More nested conditions...
}

Agentic Approach

// New Way: Intelligent Agent
const agent = new ReturnAgent({
  context: customerHistory,
  policies: brandGuidelines,
  learning: historicalOutcomes
});

const decision = await agent.evaluate(return);
// Agent considers context, learns from outcomes,
// and makes nuanced decisions automatically

What Changes with Agentic Ops?

This feels less like brittle automation and more like a capable operator collaborating with your stack, 24/7.

Learning loops, not rule sprawl

Every interaction feeds intelligence back into the system, making processes simpler and smarter over time.

Contextual decisions

Agents understand why a task exists, not just what to click, so exceptions don't require more rules.

Outcome‑first measurement

Success isn't "tasks completed"; it's results delivered (saves, resolutions, fulfilled promises).

Governance by design

Human‑in‑the‑loop controls, audit trails, and policy guardrails reduce brand risk and keep actions explainable.

eCommerce vs iCommerce: The Paradigm Shift

Aspect
eCommerce (Old Way)
iCommerce (New Way)
Core PhilosophyRule-based automationLearning-driven adaptation
Decision MakingIf-this-then-that logicContextual reasoning with AI
Exception HandlingMore rules & manual fixesSelf-improving responses
ScalabilityLinear cost increaseLogarithmic efficiency gains
Knowledge BaseStatic documentationDynamic learning loops

"The future of commerce isn't about more automation—it's about smarter automation that learns from every interaction and improves continuously."

Dominic Steil, StateSet CEO

What It Looks Like in the Wild

Subscriber Retention

Recover churn-risk subscribers with personalized save actions at the perfect moment.

Churn reduction43%
Save rate improvement2.8x
Response time<30 sec

Warranty Processing

Approve and ship warranty replacements automatically within policy thresholds.

Auto-approval rate89%
Processing time2 min
Customer satisfaction+38%

Inventory Optimization

Reroute orders proactively when inventory risks emerge across marketplaces.

Stockout prevention92%
Inventory turns+45%
Fulfillment speed1.5 days

Fraud Prevention

Flag high-risk orders for review with evidence automatically pre-compiled.

Fraud detection96%
False positives-73%
Review time5 min

The Power of Compound Learning

Unlike traditional automation that degrades over time, agentic systems improve with every interaction. Each decision becomes training data, making tomorrow's operations smarter than today's.

Adoption Made Simple

We hear it all the time: "AI Agents sound powerful and complicated." Our promise: we make complicated → simple; we make scary → not scary through our agentic flywheel.

How we de‑risk adoption:

Guided rollout with staged automation (observe → suggest → auto‑act with approvals).
Policy guardrails and brand attributes you control.
Auditability with reasoning traces and full action history.
Backtests & evals on your historical data before going live.
Predictable pricing so finance knows exactly what success costs.

On the Future of Intelligent Commerce: Cohesion Is Key

We're in the midst of a wave of generative and agentic technologies that let businesses scale faster than ever:

Autonomous AI agents
Complete CX and operations workflows end‑to‑end
Image generation
Create and refine creative assets instantly
Voice agents
Converse and take actions in real time
Computer‑use agents
Operate browsers/apps hands‑free

What's missing isn't capability; it's cohesion. We need a single intelligent commerce suite with a battle‑tested orchestration engine that coordinates these modalities safely, predictably, and at scale.

That's the future we're building at StateSet:

  • Intelligent reporting in seconds across all commerce data.
  • Order orchestration across Shopify, NetSuite, and your 3PLs.
  • CX agents that handle order/subscription changes, replacements, and returns.
  • Creative & social agents that produce on‑brand assets and publish to the right channels.

StateSet iCommerce orchestrates the operational stack end‑to‑end so your teams can focus on strategy, brand, and growth.

Our Thesis

The future of intelligent commerce is an agentic workflow orchestration engine spanning the entire business.

The Journey So Far: Why I'm Building StateSet iCommerce

My career has been shaped by two threads—language intelligence and distributed systems—leading directly to the founding of StateSet.

2013
Language & Systems
2015
Early Agentic Thinking
2019
Founding StateSet
2021
Generative Inflection
2026
Category Creation

Key Milestones

  • 2013 — Foundations in Language & Systems. Computational Linguistics (Prolog grammar parsers for English↔Japanese) alongside Raspberry Pi–driven ASIC Bitcoin miners.
  • 2015 — Early Agentic Thinking at Apttus. Sales Engineer MVP and developed Max, a Quote‑to‑Cash agent in Slack that generated contracts and built quotes.
  • 2015–2019 — Blockchain Builder. Deep work on Ethereum, stablecoins, and agentic apps; Consensus Hackathon winner three years running with awards from Microsoft, Wanxiang Blockchain, IBM, and others.
  • 2019 — Founding StateSet. Applied those learnings to the emerging DTC market, helping scale our first customer from ~$5M → $100M GMV. The systems became the cornerstone of our iCommerce Engine.
  • 2021 – 2025 — Generative Inflection. With the GPT wave, we proved that LLMs could power CX and, eventually, all of commerce ops. We expanded into computer‑use automation and voice agents.
  • 2026 – 2030 — Category Creation and Market Dominance. We've coined iCommerce: a category‑defining system for how modern brands will operate, drawing on over a decade of applying these technologies across Salesforce and Shopify.

Why It Matters for Growth

Reliability at scale

Fewer broken flows, fewer post‑mortems.

Operational leverage

Teams move from maintaining bots to managing outcomes.

Brand safety

Clear guardrails and explainability.

Cost clarity

Predictable operating costs beat per‑automation surprises.

How StateSet Makes It Real

At StateSet, AI‑native agents power the iCommerce Engine, connecting Shopify, Amazon, TikTok Shop, NetSuite, 3PLs, and more to deliver outcomes, not just actions.

Service‑as‑Software

Autonomous agents delivering measurable results.

Outcomes based pricing

No per‑ticket/per‑conversation fees.

Fast time‑to‑value

Two‑week implementation, no annual lock‑in.

Governed ops

Human approval paths, auditability, and policy controls (SOC 2).

Vision

Our goal is to partner with the top 5,000 Shopify brands for enterprise‑grade, agentic operations, plus hands‑on guidance on the latest models, modalities, and best practices to help you scale past $100M+.

Our team is committed to bringing the next generation of intelligent commerce to operators worldwide so you can move faster, de‑risk growth, and compound outcomes.

"Ready to transform your commerce operations? The shift from eCommerce to iCommerce isn't just about technology—it's about fundamentally reimagining how your business operates, learns, and scales."

Ready to Join the iCommerce Revolution?

See how StateSet can transform your operations in just 2 weeks.