Panayiotis AlexandrouHands-on CTO · VP / Head of Engineering

Selected work

Systems I designed and built

Two of these are live production systems with real customers on them, built outside my day role. The third is the platform I own at work, described by its architecture rather than by its owner.

Notevex

ERP-grade multi-tenant SaaS

A distributed multi-tenant SaaS unifying CRM and sales, operations, a developer API and a client contact portal, across responsive web and native mobile.

  • Commercially validated — paying customers run their day-to-day operations on it today, acquired entirely by word of mouth with no marketing spend.
  • Enterprise security and RBAC: JWT with certificate validation, 2FA, full tenant isolation, granular per-user CRUD permissions, and a complete audit trail.
  • Real-time updates via SignalR, and a REST API with HMAC-SHA256 signed webhooks.

.NET 10 · MassTransit/RabbitMQ · Redis · PostgreSQL · Angular · React · Flutter · Firebase — ~10 services containerised on Kubernetes with auto-scaling and full observability.

notevex.com

Operavex

Operations, analytics and invoicing platform

A bespoke distributed platform running a business's day-to-day operations, analytics and invoicing end to end.

  • Real-time dashboards turning operational data into decisions; invoicing generated directly from live operational data.
  • Shares Notevex's event-driven architecture, giving it a proven path to a future multi-tenant SaaS offering.

.NET 10 · MassTransit/RabbitMQ · Redis · PostgreSQL · Vue.js — ~10 services on Kubernetes with auto-scaling and full monitoring.

AML/KYC platform

Architecture highlights

The regulated-fintech platform I own at work, described by its architecture. Transaction and profile monitoring, client onboarding, risk assessment and sanctions/PEP screening.

  • Transaction and profile monitoring — ~30 microservices designed and delivered start to finish: gateway APIs, MCP servers, and the consumers and processors behind a complete monitoring solution.
  • Client onboarding — ~28 microservices across a gateway layer (REST/gRPC and reporting APIs) and single-responsibility processors coordinated by SAGA orchestration; broker-absorbed back-pressure replaced dropped requests from partners without correct rate limiting or throttling.
  • In flight — risk assessment being extracted from the monolith onto the same architecture; sanctions and PEP transaction screening fully architected and queued for delivery.

Product suite and other systems

Independent and client work

The same event-driven platform extended into a vertical suite in active development.

  • Hotelvex (hotel management), Payrovex (payroll), Accountvex (accounting) and Eventvex (event management and booking).
  • 10+ products built independently across web and mobile — 3 running in production today, 5+ live including marketing sites — plus automation bots and MCP servers.

Build principles

The same five decisions turn up in everything below, and they are deliberate rather than habitual.

  1. Event-driven, broker-first messaging

  2. Single-responsibility services

  3. SAGA orchestration for distributed consistency

  4. Containerised and autoscaled by default

  5. Observability from day one

A note on the independent work

Notevex and Operavex are live production systems I built and operate independently, outside my day role. A walkthrough or an architecture deep-dive of any system on this page can be arranged on request.

Want the detail behind any of this?

A walkthrough or an architecture deep-dive of any system on this site can be arranged on request.