four30.co Book a conversation
All projects

Startup

WDDL

An agentic retail operations platform for independent stores. It reads sales and staffing data and shows managers what needs attention today.

Type
Startup
Year
2025
Our role
Co-founder and CEO
Stack
Next.js, React, TypeScript, Supabase, Vercel, Claude
WDDL: An agentic retail operations platform for independent stores. It reads sales and staffing data and shows managers what needs attention today.

WDDL (Workforce Data Decision Layer) is an agentic retail operations platform for independent stores. Tony Blount is its co-founder and CEO. He started the codebase in November 2025, and the first production deploy went out on December 17, 2025. Marcus Johnson-Smith, a multi-store owner-operator, was WDDL's first customer before he joined as co-founder and CRO.

Why it exists

Larger chains have district managers, analysts and workforce software. An independent retailer has spreadsheets and the owner's evenings. The pitch deck lists what that looks like on the floor: schedules built in Excel and retyped into payroll, managers doing admin with less time to coach, and owners who find out after the money is gone. WDDL is built to fill that missing operations role in every store.

What it does

The agent opens the day. The Today screen shows sales so far, transactions, average basket and labor, then a list of what needs a manager: an associate whose revenue has dropped, a notice still waiting for acknowledgement, the tasks left open.

The Today screen for a demo store: sales, transactions, average basket and labor across the top, then a Needs You list led by a coaching alert.
Today, on a demo store with fictional staff and sales.

Around it sit the tools a store runs on: scheduling with expected demand and time-off requests, team chat and push-to-talk radio, tasks, announcements that staff acknowledge, training with quizzes, inventory, incident logs and recognition. An owner with several stores switches between them from one login.

A week of shifts by person, under a strip that rates each day's expected demand from light to busy.
The schedule puts expected demand above the week's shifts.

How it works

WDDL pulls each store's point-of-sale data at set times through the day. From it, the platform writes coaching at five levels, daily through yearly, for the store and for each sales associate. Drafts land in a review queue, where a manager refines or approves each one.

The Coaching Review queue: four daily store drafts, each with Refine and Approve buttons.
AI-generated coaching waits for a manager's approval.

An AI chat, built on Anthropic's Claude, answers questions about the store's own numbers, such as which associate sold the most yesterday.

Under the screens sits one data model for the whole store. Each store is a tenant in Supabase Postgres, and row-level security keeps each person's reads inside the stores they belong to. The app is Next.js on Vercel, with Vercel cron jobs running the syncs and scheduled reports.

A deck slide titled One Model the Whole Store: Square for commerce, Homebase for workforce, and WDDL for operations, one shared context across store tools and outside agents.
From the pitch deck: where WDDL sits beside commerce and workforce platforms.

Where it stands

WDDL is live at wddl.app. The pitch deck reports an early paid beta, live and paid in two stores since July 1, 2026. The platform is built to deploy across verticals: a demo group of four fictional fashion boutiques runs on the same code, with labels for that trade. The plan is retail first, then other deskless businesses such as auto repair, restaurants, vet clinics and home services. Mobile apps for floor teams are next on the deck's list.