Adopt

Adopt is a cosmetics brand whose B2B order management relied entirely on manual Excel-based workflows.

Orders, client data, and operational logic were handled through spreadsheets that were fragile and error-prone, difficult to scale or audit, and highly dependent on individual knowledge.

When I joined the project, the situation was already critical: the project was severely behind schedule, expectations were high, and pressure to deliver was immediate.

A digital replacement was urgently needed — not as a long-term experiment, but as a production system that had to work quickly and reliably.

The main challenge was delivering a reliable system under significant time pressure, without sacrificing correctness or long-term maintainability.

The existing Excel workflows encoded essential business logic implicitly. Replacing them meant understanding real operational behavior quickly, formalizing it into a structured data model, and building a system that teams could trust immediately.

All of this had to happen while recovering from existing delays and keeping the delivery timeline intact.

I joined as Tech Lead and Full-Stack Developer, taking responsibility for both technical direction and delivery.

My responsibilities included:

• Leading development across the stack
• Designing the data model mapping Adopt's requirements to Commercetools
• Delivering core features on the Next.js B2B web platform and its BFF
• Engineering a custom import system to handle large order datasets
• Coordinating with product owners, Adopt stakeholders, and the Commercetools support team

Given the project state on arrival, I had to make fast, pragmatic decisions while keeping the system coherent and extensible.

1. Building on an existing platform without re-litigating decisions

Commercetools had already been selected through a proof of concept before I joined.

Rather than revisiting this choice under pressure, I focused on understanding its constraints quickly, using its strengths effectively, and designing around its limitations where needed.

This allowed the team to regain momentum instead of losing time on architectural debates.

2. Formalizing spreadsheet logic under time pressure

The Excel workflows contained years of accumulated business rules.

I worked with stakeholders to extract implicit logic, clarify edge cases, and identify which behaviors were critical vs incidental.

This ensured the system preserved essential behavior while eliminating many sources of error.

3. Designing a scalable data model in Commercetools

I defined the full data structure using Commercetools' schema and custom types, mapping Adopt's B2B requirements into a model that was explicit, validated, and scalable.

This became the backbone of the new platform.

4. Engineering a custom import system

The most technically challenging part of the project was the custom import system, built from scratch with Node.js on top of the Commercetools import API.

It enabled reliable ingestion of large order datasets, recovery from partial failures, and predictable processing under load.

This component was critical to moving away from spreadsheets without data loss.

Despite the project being significantly behind schedule when I joined:

• The platform was delivered on time
• Fragile Excel workflows were replaced with a robust digital system
• B2B order management became faster, safer, and more reliable
• Operational errors caused by manual handling were significantly reduced

The client expressed strong satisfaction with the result, highlighting substantial time savings.

The project's success also led to a partnership opportunity with Commercetools, where our team later built a white-label product to accelerate onboarding for future clients.

• Commerce platform: Commercetools
• Web: Next.js (B2B platform + BFF)
• Backend tooling: Node.js (custom import system)