Beluga Analytics: retail intelligence validated by a multinational.
- Product
- IoT and retail analytics SaaS platform
- Pilot
- Consumer-goods multinational
- My role
- Co-founder and sole developer
- Period
- Oct 2024 – Dec 2025 (paused pending hardware)
The challenge
Beluga needed a platform that could take in high volumes of retail data, coming from IoT devices and store operations, and turn it into reports a manager understands without building spreadsheets. And it had to be ready for evaluation by a consumer-goods multinational in a real pilot.
What was built
- A complete platform from scratchA .NET SaaS with Clean Architecture and an Angular 17 dashboard, designed to grow without being rewritten.
- Retail data at scaleGoogle BigQuery routines that process large volumes of retail data without slowing operations down.
- Reports that write themselvesMachine learning models on Vertex AI generate the reports automatically, with no manual work.
- Security and continuous deliveryContainers on Cloud Run with routing, DNS and CI/CD, JWT authentication and credentials protected in Secret Manager.
Technical details
See the engineering pageBackend
.NET with Clean Architecture, REST APIs and JWT authentication.
Frontend
Angular 17 with RxJS and Signals for the analytics dashboard.
Data and AI
BigQuery pipelines and predictive reports with Vertex AI models.
Infrastructure
Docker on Cloud Run, routing and DNS, CI/CD and secrets in Secret Manager.