Montassar
Fatfouti

Design, development and evolution of a SaaS internal communications and employee experience platform (Employee Hub), covering the full software lifecycle — architecture, technical design, backend features, performance optimisation, and third-party integrations.

• Led technical and functional analysis across multiple product streams — translating business requirements into system designs, API contracts, and infrastructure blueprints
• Made key architectural decisions on service boundaries, data models, event flows, and deployment topology; authored specs that guided cross-functional engineering teams
• Designed and developed backend microservices (TypeScript, Python, Flask) and REST / API Gateway layers serving millions of daily users
• Set up, optimised and supervised CI/CD pipelines with GitHub Actions, Jenkins, and ArgoCD
• Automated deployments on GKE and managed GCP infrastructure (Cloud Run, Pub/Sub, Cloud SQL)
• Implemented KEDA-based event-driven autoscaling on GKE, using Redis queue depth as scale triggers — significantly reducing idle compute cost
• Designed Redis caching and session layers (Cluster + Pub/Sub) to handle traffic spikes with low-latency horizontal scaling
• Owned on-call rotations; drove monitoring, alerting, and observability with Datadog
• Mentored engineers, conducted architecture reviews, and enforced best practices across the backend guild

Built an automation platform for the telecom / fibre-optic industry to centralise and automate manual workflows (FME, ETL pipelines) and validate network node configurations through API-driven controls.

• Conducted the full technical and functional analysis — mapping manual field processes to system components and defining the data model, API surface, and integration points
• Made all architectural decisions solo: technology selection, service structure, and API design strategy
• Designed and delivered a Django-based platform that replaced manual workflows with automated API-driven validations
• Exposed REST APIs consumed by field teams for real-time optical node configuration checks

Automated and maintained an insurance contract subscription system built on Blockchain Quorum, serverless functions, and Databricks clusters for price visualisation and prediction.

• Maintained a Blockchain Quorum-based system for insurance contracts and quotes
• Built serverless event-processing functions and automated the contract quote pipeline
• Used Databricks clusters for price visualisation and ML-based prediction pipelines

Contributed to Expensya — a SaaS expense management platform serving 5,000+ customers across 192 countries (acquired by Medius in 2023) — building core application features and the intelligent OCR+ pipeline that automatically digitises paper receipts and expense notes.

• Built and maintained CRUD features across the core expense management application: expense reports, approval workflows, user management, and multi-currency handling
• Contributed to OCR+, the product's core intelligence — a pipeline that automatically digitises paper receipts by extracting key fields (amount, date, merchant, VAT, currency) from raw scanned images
• Implemented classification and categorisation modules using TF-IDF vectorisation, regex pattern matching, and keyword-based scoring to auto-assign expense categories (meals, transport, accommodation) per company policy
• Combined multiple NLP techniques — optical character recognition, information retrieval, statistical text classification, and structured data extraction — into a single robust and extensible pipeline
• Worked across the full Python / C# / ASP.NET stack on Azure, contributing to both backend services and the data extraction layer