 🚀 Learnova |
AI-Powered EUEE Prep Platform — Production system at Solanova Technologies. NestJS/TypeORM backend with Elasticsearch for curriculum search, RabbitMQ for async content processing, MinIO for asset storage. Frontend on Vite + React 18 + shadcn/ui + TanStack Query, tested with Vitest + Playwright. 
|
| 👤 Name | Ewnetu Tesfaye ("Kiyaeh") | 📍 Location | Addis Ababa, Ethiopia |
| 💼 Role | Full-Stack Engineer, Solanova Technologies | 🎓 Education | B.Sc. Computer Science, Samara University |
| 🛠️ Started Coding | 2021, with C++ | 🌍 Long-Term Goal | AI/ML development in Ethiopia & Africa |
| 🔭 Currently On | Learnova — EUEE AI prep platform | 🎯 Interests | Clean arch, distributed systems, fitness, good food |
I'm a full-stack engineer who likes solving the same problem more than once — on purpose. Most of my recent work traces one question across three different architectures: how do you build an EUEE prep platform that scales? Each answer taught me something the last one couldn't.
Currently deep in NestJS + Elasticsearch + RabbitMQ for production, experimenting with Spring Cloud microservices on the side, and keeping MERN + React Native roots active through client work.
flowchart LR
subgraph P1["📚 Phase 1 — Final Year Project"]
A["🎓 EUEE Prep Platform\nNext.js + NestJS + Prisma\nAWS EC2 · Monolith-leaning"]
end
subgraph P2["⚡ Phase 2 — Production (Active)"]
B["🚀 Learnova\nNestJS + TypeORM\nElasticsearch · RabbitMQ · MinIO\nSearch/async-driven"]
end
subgraph P3["🔬 Phase 3 — Experimental"]
C["🧪 Automate Tech\nSpring Cloud · Eureka\nAPI Gateway · agentbot\nTrue microservices"]
end
P1 --> P2 --> P3
style P1 fill:#1e3c72,stroke:#2DD4BF,color:#fff
style P2 fill:#0f9b8e,stroke:#2DD4BF,color:#fff
style P3 fill:#134e4a,stroke:#2DD4BF,color:#fff
Same domain. Three deliberate architectural philosophies. Monolith-leaning → Search/async-driven → True microservices.
| Project | Description | Stack |
|---|---|---|
🚀 Learnova
|
AI-powered EUEE prep platform. Elasticsearch curriculum search, RabbitMQ async processing, MinIO assets. TanStack Query frontend. |
|
🧪 Automate Tech
|
Java 21 / Spring Boot 3 multi-module system. Eureka service discovery, API Gateway, isolated agentbot AI module — independently deployable. |
|
💬 Real-Time Chat
|
7-service microservices system. Splits message persistence from WebSocket transport for independent scaling. Docker Compose orchestrated. |
|
|
🐟 Tommy Fish ERP
|
Production ERP for a Dire Dawa distribution business. Tiered pricing, stock deduction, branch transfers, payroll, full reporting. |
|
|
🍽️ Multi-Tenant Restaurant
|
Slug-based multi-tenancy serving multiple restaurant clients from one codebase. 4 role-scoped UIs: Customer, Cashier, Kitchen, Admin. |
|
|
🏝️ Afar Tourism Bureau
|
Government client delivery. MERN CMS + admin dashboard for regional tourism bureau. Deployed on Heroku with AWS S3 media storage. |
|
| System Design Patterns |
|
| Spring Cloud Depth |
|
| Elasticsearch Internals |
|
| AI/ML Engineering |
|
I'm looking to work on:
Fitness and good food keep the mind as sharp as the code — you'll usually find me either debugging a service mesh or at the gym, rarely anything in between. Always up for a conversation about clean architecture, AI in emerging markets, or the best place to eat in Addis.
