Scott Lee

Laboté is a skincare startup specializing in fully customized skincare products. Its value proposition is that each skincare product, such as day cream, night cream, and serum, is unique and based on a skin diagnostic. Users onboard and respond to a questionnaire, which is processed into a skin profile that’s validated by a licensed pharmacist. Product formulas are tweaked to match a particular skin profile with a one-to-many relationship between a user and their skin profile.

I created the system architecture, hired a UX designer, and built a distributed team on less than 50,000 EUR. In addition to creating significant complexity in terms of SKUs and product history, fully traceable records were required, as skincare is regulated by food and drug authorities. A custom platform was required to piece all these systems together, and at least 60% was new code.

One of the three major luxury groups was facing challenges in its South Korean market strategy with KakaoTalk, the Korean equivalent of WeChat and the primary internet platform used locally. Brands could not integrate their eCommerce, CRM, and ERP systems with KakaoTalk, leading to decreased user engagement and marketing challenges.

Since the luxury brands used a mishmash of different providers, the project required a feasibility study across the brand-specific systems and KakaoTalk’s own APIs. The result was a comprehensive implementation guide on the API level in conjunction with a Salesforce-based middleware layer.

The client was a hybrid, AI-based marketplace matching company for tech professionals. The premise was to use AI to provide a first matching, screening, and qualification and then use recruiters to provide the soft touch and coaching required to close the deal.

The platform was originally based on a white-label solution proposed by an American startup specializing in AI tech for recruitment. This startup eventually built a heavily modified and effectively forked version of their platform for the client.

The project was to take over this platform and carry out very necessary UX improvements, internalize development with in-house resources, and partially integrate with the corporate parent group’s systems for additional synergies. This last part required significant change and stakeholder management, especially as the startup was financed by the parent group through a traditional budget process.

The project was discontinued due to the aftereffects of Brexit and the COVID-19 pandemic.

Cheerz used Stripe across its major markets for card payments, with France and Italy being two crucial markets. Unfortunately, due to some obscure, deeply embedded configuration issues, 2-3% of all payments failed at the very end of the conversion. These failures were caused by problems between the PSP, acquiring bank, and issuing bank, with certain details lost in transit.

The project involved detailed data analysis of payment records in Stripe’s own logs and the back office, requiring cross-matching to verify and sample payment issues. These issues were flagged, confirmed in a discussion with Stripe in which their product team admitted this bug, and a new plan of action proposed the use of ProcessOut, an alternative PSP, as a backup in case of failed payments.

I prepared and presented complete specifications and a roadmap before the handoff and my departure.