Technical Lead, NoSQL data architecture, and developer team coordination.
BackMy role Challenge Observed outcome
About confidentiality Next Project
Social Intelligence & Influencer Management Hub
An in-house platform to centralize influencer metrics and campaigns, replacing high-cost SaaS solutions with a serverless architecture on Firebase.
Confidential case
AngularFirebase FirestoreCloud Functions (Node.js)OAuth 2.0 (Multi-provider)Algolia (Search)
Migrating operations from multiple paid platforms (e.g., Squid) to a proprietary system that required normalizing heterogeneous data from Instagram, TikTok, X, and YouTube in real-time.
The v1 delivery centralized campaign lifecycles with automatic metric harvesting via OAuth, eliminating third-party software costs and ensuring data accuracy for influencer filtering.
Concrete signals from this case
- Led development from inception to a stable v1, coordinating a team of developers and a dedicated UX designer.
- Architected a flexible data layer using Firebase Firestore (NoSQL) to handle schema variations across different social networks.
- Implemented a multi-channel OAuth orchestrator, allowing influencers to connect and manage permissions for multiple social accounts simultaneously.
- Built manual matching filters based on engagement and real reach, powered by periodic metric update routines.
I developed this project to solve a growing operational cost for the client: reliance on external social intelligence platforms. The goal was to build a proprietary tool that centralized the entire campaign lifecycle—from influencer onboarding to final metric analysis.
Architecture and Data Flexibility
One of the biggest technical challenges was data normalization. Every social network (Instagram, TikTok, YouTube, X) delivers metrics and formats differently. We chose Firebase Firestore for its NoSQL nature, allowing the system to be dynamic enough to store channel-specific attributes without rigid database constraints.
OAuth Orchestration and Onboarding
The influencer onboarding flow was the most complex part of the interface. We built a centralized hub where users could authenticate their socials via OAuth in a continuous flow. These permissions ensured the platform could perform periodic data harvests (via Cloud Functions), keeping the influencer database updated with followers, engagement rates, and content categories.
Leadership and Collaboration
In this project, I acted as the technical lead, mentoring a team of developers (mostly former students) and working alongside a dedicated UX designer. My role covered everything from tech stack decisions to implementing complex business logic in Firebase Functions, ensuring campaigns followed strict briefings and automated schedules.
Visual references
Below are diagrams and screenshots showing the system in operation:
Campaign data and business secrets have been anonymized. The description focuses on technical leadership and infrastructure/product decisions.