Case Study

Forward-Deployed AI Operations In A Public Company

A public-safe view of how agentic workflows can support supply chain, ecommerce, CRM, ERP, finance, reporting, and executive operations inside a lean operating company.

Joaquin Abondano - COO, Innovative Eyewear / Lucyd (NASDAQ: LUCY)

Operating Layer Sales, marketing, fulfillment, supply chain, finance, and executive operations.
Systems Context ERP, CRM, ecommerce, marketplace, reporting, and communication workflows.
Governance Human approval layers, escalation paths, and executive accountability.

Context

Modern operating companies run across fragmented systems: ERP, CRM, ecommerce, marketplaces, finance tools, vendor communication, logistics updates, product launch plans, and executive reporting. The constraint is rarely access to another AI demo. The constraint is deployment ownership: deciding where AI belongs, how it is supervised, and how it improves daily execution without creating uncontrolled risk.

The Operating Problem

The operating challenge is coordination. Teams need faster visibility across sales, fulfillment, inventory, vendors, marketing, finance, and customer channels, while keeping consequential decisions in human hands.

Deployment Approach

The approach is to deploy AI as an operating layer, not as a replacement for judgment. Agents monitor workflows, summarize changes, draft routine outputs, flag exceptions, and route decisions back to accountable humans.

Example Workflows

Sales And CRM: Pipeline summaries, account follow-up prompts, stale-deal alerts, and executive sales snapshots.

Supply Chain And Fulfillment: Vendor follow-up drafts, fulfillment status monitoring, inventory exception alerts, and logistics summaries.

Ecommerce And Marketplace Operations: Product listing checks, campaign summaries, customer feedback themes, and channel performance monitoring.

Finance And Reporting: Variance prompts, cash-flow visibility support, weekly operating summaries, and executive briefing preparation.

Executive Operations: Daily priorities, open-loop tracking, decision logs, meeting prep, and cross-functional follow-through.

Operating Principles

The goal is controlled leverage. AI should reduce manual drag, increase visibility, and improve follow-through while preserving accountability for material decisions.

Results Framing

This operating model creates more leverage per person by turning scattered workflows into visible, reviewable, and repeatable execution loops. It helps leadership teams move faster without pretending that AI removes the need for judgment.

What Stays Private

This page intentionally excludes confidential vendors, margins, purchase order quantities, internal dashboards, board materials, compensation details, runway assumptions, private financial interpretation, and sensitive workflows. The point is to show the deployment pattern without exposing the operating system of the company.

Where This Fits

This is the practical layer behind forward-deployed AI operations: mapping business workflows, connecting systems, designing agent responsibilities, governing write actions, and keeping executive judgment where it belongs. It is the work between AI strategy and production reality.

Read the JBOT Protocol methodology ->