Guillaume Kpomda

Lead fullstack developer on an internal dashboard developed for Samshield, a B2B platform dedicated to the configuration and management of customizable helmets. The application allowed internal teams to manage product models, configurations, pricing rules, and resellers, while providing analytics and activity tracking tools.

I designed and developed the interface with Next.js (App Router), React, and TypeScript, using Apollo Client to consume a GraphQL API. The frontend architecture was based on a domain-driven organization and automatic generation of typed hooks via GraphQL Code Generator. Particular attention was paid to managing complex interfaces related to product configuration (models, parts, option families, colors, pricing rules) as well as optimizing performance when handling large volumes of data.

On the backend, I participated in the development of a GraphQL API built with Node.js, GraphQL Yoga, and Pothos, connected to PostgreSQL via Prisma ORM. The system was based on rich business logic around hierarchical product configuration and dynamic pricing rules. Redis was used for caching, event management via pub/sub, and processing analytical flows, while an analytics pipeline tracked sessions, user interactions, and usage metrics.

The entire platform was containerized with Docker and deployed via a CI/CD pipeline, enabling reliable deployments and continuous application evolution. The objective was to provide a robust and performant tool for managing complex product configurations and facilitating operational and analytical oversight of the platform.

Lead fullstack developer at Aara, a concierge application augmented by artificial intelligence, creating AI agents dedicated to the world of hospitality and concierge services.

I design and develop AI agents to assist concierges in their daily tasks and facilitate information access for clients. The architecture relies on Django and Pydantic AI, with direct integration on WhatsApp via Twilio and Meta APIs. I use a vector database (Pinecone) to manage a local address recommendation system and a RAG (Retrieval Augmented Generation) mechanism to provide reliable and contextualized information to guests.

The goal is to implement practical and efficient tools capable of effectively supporting the work of concierges and improving the quality of service provided to clients.

Lead Mobile Developer at Docent, an art discovery mobile application using artificial intelligence to recommend personalized artworks to users.

Mobile Development: Implementation and improvement of application features under React Native, React.js, and TypeScript, in collaboration with the product team, designers, and backend developers.

Performance Optimization: Work on the fluidity of the user experience with optimized transitions and animations for a smooth and responsive rendering.

Backend & API: Collaboration with backend developers for the integration of Node.js and GraphQL, ensuring fast and efficient communication between front and back ends.

Agile Methodology: Work in Scrum mode with rapid iterations, regular reviews, and strong interaction with business teams.

User Experience: Improvement of interactions and navigation to offer an intuitive and pleasant interface.

Involvement in the entire mobile and web part of the project with a strong focus on performance, UX/UI optimization, and JavaScript best practices.