Soum operates a multi-category marketplace across the MENA region with a polyglot backend (Node.js, Go, Python) deployed on Kubernetes across AWS, GCP, and Azure. The tech stack reveals heavy infrastructure investment: ClickHouse for analytics, TensorFlow/PyTorch for ML, and a full observability layer (New Relic, Prometheus, Grafana). Active projects signal aggressive scaling—recommendation systems, fraud detection, demand forecasting, and a predictive modeling platform are live, while pain points (low-latency scalability, data pipeline bottlenecks, real-time data flows) suggest they're hitting growth limits. The hiring mix is data-heavy (13 roles) relative to engineering (11), pointing to a company transitioning from product velocity to analytics-driven decision-making.
Tech Stack 58 technologies
What Soum Is Building
◆Challenges
- High availability disaster recovery
- Low-latency system scalability
- Scaling data pipelines
- Process gaps and automation
- Tax compliance coordination
- Handling large-scale real-time data flows
- Automation improvements
- Ensuring data integrity
- Mission-critical system stability
- Analytics adoption across teams
▲Active Projects
- Recommendation systems
- Develop accounting documentation, dashboards, and sops for scale
- Secure infrastructure implementation
- Scalable data pipeline architecture
- Demand forecasting
- Data infrastructure for analytics and ml models
- Fp&a initiatives and strategic modeling
- Ci/cd pipeline enhancement
- Predictive modeling platform
- Fraud detection
