Mizan a fitness and meal tracking application with AI and MCP integration.
- Developer
Hello! I’m Euael M. Eshete, a software engineer based in Addis Ababa, Ethiopia, focused on building secure, scalable backend systems and AI-enabled platforms. I enjoy turning complex requirements into clean architectures, bootstrapping cloud-native services, and delivering practical solutions that balance performance with reliability.
I also mentor students and contribute to community initiatives, sharing knowledge about full-stack development, AI/LLM integration, and cloud technologies to help others grow in emerging markets.
Mizan a fitness and meal tracking application with AI and MCP integration.
Security Architecture: Designed and implemented the RBAC (Role-Based Access Control) module for the client-facing security dashboard using ASP.NET Core, ensuring strict data isolation between tenants. Containerization Strategy: standardized the deployment pipeline by containerising legacy .NET services with Docker, reducing environment parity issues between dev and prod. Front-end Performance: Optimised Razor Pages rendering logic to reduce Time-to-Interactive (TTI) for high-density data grids.
Mentored over 150 students in full-stack development, AI/LLM integration, and cloud technologies, fostering persistence and adaptability in emerging market contexts.
High-Concurrency Architecture: Architected a non-blocking, asynchronous backend using FastAPI (Python) and Litestar, scaling from 1,250 to 58,000 monthly messages while supporting 2,000 active users. Cost-Efficient Migration: Engineered a zero-downtime storage migration moving media assets to GCP Cloud Storage, rewriting asset delivery to reduce infrastructure overhead by 94%. LLM Pipeline Engineering: Built a Retrieval-Augmented Generation (RAG) pipeline, integrating vector databases and OpenAI APIs; optimized prompt caching strategies to reduce API latency and token costs, increasing user retention by 30%. Infrastructure Scaling: Orchestrated the transition from monolithic deployment to containerised services on GCP Cloud Run, implementing auto-scaling policies based on CPU and latency metrics. Market-Specific Engineering: Developed a lightweight Telegram Mini-App wrapper, addressing authentication and session management challenges within the Telegram webview environment, onboarding
Post a job and invite Euael, or contact them directly. No hidden commission.