Eight weeks to go-live, growing together ever since.

The brief

An American gaming peripheral manufacturer and distributor came to us in May 2025. Multi-channel, multi-warehouse, growing fast. Their on-premises legacy ERP had served them well for years but couldn't keep up: data was siloed across teams, reporting lagged behind reality, warehouse picks were error-prone, and finance close took longer every month.

The mandate: migrate to cloud Business Central, integrate the existing Magento storefront, deploy a warehouse management system that actually worked on the floor. Go-live target: July 1, 2025. That's eight weeks from contract signature.

What we shipped in eight weeks

May 2025 contract signed. June was training, data migration, and customization in parallel — chart of accounts mapped, master data cleaned, open transactions migrated, role centers tuned per team, custom posting rules and shipping documents built. July 1, 2025: hard cutover. Finance, ecommerce, and warehouse all moved to the new stack on the same Monday morning.

Three platforms launched the same day:

• Microsoft Dynamics 365 Business Central — full migration from the legacy on-premises ERP, with custom AL extensions handling client-specific workflows
• Magento ↔ BC integration — bi-directional sync for orders, customers, fulfillment status, and inventory, with custom logic for the edge cases native connectors couldn't address
• WMSPilot deployment — our Android-based warehouse management system, running native against the BC API. Pick, pack, receive, cycle count, putaway — all on handheld scanners from day one

Why WMSPilot exists

The off-the-shelf WMS options in the Business Central ecosystem fall into two buckets: enterprise-priced products built for distribution centers ten times this client's size, or lightweight scanner apps that don't really integrate with BC business logic. Neither fit.

So we built our own. WMSPilot is an Android-native handheld application that talks directly to BC via REST APIs. It handles pick, pack, receive, cycle count, putaway, and inter-warehouse transfer workflows — and crucially, it respects BC's posting rules, dimensions, and item tracking codes natively rather than as an afterthought.

Built first for this client. Refined daily through real warehouse use since July 2025. Now a productized Darbtek offering serving multiple distributors across packaging, electronics, textiles, and hardware. The fact that working warehouse staff use it on every shift — and feed back what's slow, what's broken, what's missing — is what makes it different from a stock scanner app.

The hard parts: July to September 2025

Any ERP go-live has a tail. The first thirty days you find the bugs you missed. The next thirty you find the workflow assumptions you got wrong. The next thirty you find the edge cases nobody mentioned in discovery.

For this client, July to September 2025 was that tail. Pickers on the floor told us which sequences felt slow, and we re-sequenced. Posting rule edge cases surfaced for special-order items, and we patched. Magento integration timing issues appeared during high-volume promotions, and we re-tuned the throttling. We adjusted, deployed, repeated.

By end of September the system was stable, the team was confident, and finance close was back to taking days instead of weeks.

We don't pretend the first 90 days are smooth. We do commit to being on the line when they're not.

Platform evolution: Magento to Shopify (Nov 2025 – Apr 2026)

By late 2025 the client had outgrown Magento. The team wanted a faster admin, better B2B controls, and a simpler customization surface. We agreed.

November 2025: we started building a new Shopify storefront in parallel with the live Magento store. The work spanned five months across several workstreams running concurrently:

• Shopify storefront build — catalog mapping, theme development, B2B Company Location setup, customer migration plan
• BC ↔ Shopify integration — replacing the Magento integration layer with a Shopify-native one, while keeping WMSPilot pointed at BC throughout
• Custom Shopify app development — to fill the gaps the native Shopify B2B feature set didn't cover for this specific business
• Marketing channel reconnections — paid acquisition channels, product feeds, and tracking all rewired against the new storefront URLs and feed structure

April 2026: cutover. Magento retired, Shopify live, custom app shipped, integrations re-pointed. WMSPilot — unchanged. BC — unchanged. The back office kept running while the front of house changed shape entirely.

The "1 + AI" delivery model, applied to a long engagement

One senior consultant has owned this engagement since day one. Same person, same context, no agency handoffs, no team rotations across an entire year of work. Continuity compounds — every conversation builds on the last, every workaround stays remembered, every decision has its rationale documented.

AI handles the velocity multipliers across both the initial 8-week sprint and the multi-year evolution: generating AL code stubs, drafting integration mapping documents, computing inventory analytics across Item Ledger Entries, writing release notes and training docs. The senior owns architecture, judgment calls, and client conversations.

The result: a delivery team that scales to a four-person agency on output, costs roughly what one senior costs, and — crucially — doesn't reset its institutional memory every quarter when staff rotate.

Ongoing: quarterly inventory health

The engagement didn't end at platform launches. We run quarterly inventory health analysis against the BC Item Ledger Entry data: ABC classification by revenue contribution crossed with XYZ classification by demand variability. The output is a multi-tab Excel workbook the procurement team uses to find dead stock, identify slow movers, and prioritize reorder decisions.

Next on the roadmap: channel-split sales analysis (which SKUs perform on which sales channel), and per-SKU ROP and Safety Stock calculations factoring in real supplier lead times.

What this means for you

This client isn't a one-time engagement. They're a multi-year platform partnership that started with an aggressive 8-week sprint and kept evolving — through stabilization, through a full ecommerce platform migration, through custom application development, through ongoing inventory analytics.

If your business needs both:

• A fast initial go-live (because waiting is expensive)
• And a partner who's still around three platform changes later

— this is the playbook. Eight weeks to start. One senior + AI for the duration. Three platform transitions and counting, with nothing lost in handoffs because there were no handoffs.