Yunus A. Hassen

About

I build AI products people actually use.

I build AI products people actually use. Over the past six years I have shipped LLM-powered features to more than 100,000 monthly users, cut ML query latency by 40 percent, and lifted engagement on AI-driven interfaces by 25 percent.

Right now I lead AI and frontend work at Adilas Tech, designing RAG pipelines and agentic systems with Python, FastAPI, and vLLM underneath, and React, Next.js, and TypeScript on top.

The work I am proudest of sits closer to home: custom NLP models for hate-speech and misinformation detection in Amharic and other low-resource Ethiopian languages, where off-the-shelf models simply do not work.

Toolkit

  • Python
  • FastAPI
  • LangChain
  • LangGraph
  • RAG
  • vLLM
  • Qdrant
  • Pinecone
  • TensorFlow
  • NLP
  • Prompt Engineering
  • Gemini API
  • GPT-4o
  • Vercel AI SDK
  • React
  • Next.js
  • TypeScript
  • Tailwind CSS
  • GSAP
  • Three.js
  • Zustand
  • Node.js
  • PostgreSQL
  • MongoDB
  • Docker
  • AWS
  • GCP
  • CI/CD
  • Flutter

Based in

Addis Ababa, Ethiopia (open to remote and relocation)

Experience

Nov 2025 - Present

AI and Frontend Lead · Senior AI/ML Engineer

Adilas Tech

Leading AI development on a logistics platform: an end-to-end streaming RAG application that connects a React and SignalR frontend to a high-throughput FastAPI inference engine running vLLM, with asynchronous retrieval over Qdrant and a hybrid strategy that routes queries between self-hosted models and GPT-4o.

Nov 2024 - Jul 2025

Senior Software Engineer

SkyHigh Cloud

Led frontend development across 10+ international client projects serving 100,000+ combined monthly sessions. Integrated LLM features with the Gemini API and Vercel AI SDK, lifting user engagement by 25 percent, and cut ML query latency by 40 percent by redesigning backend data retrieval.

May 2021 - Nov 2024

Senior Software Engineer · ML Engineer

AdotSense Digital

Trained TensorFlow text-classification models for misinformation detection and built custom NLP models for hate-speech detection in Amharic and other low-resource Ethiopian languages. Also built the React platform researchers used to explore model outputs, and drove a 170 percent traffic increase through analytics-led iteration.

Dec 2019 - May 2021

Founder · Software Engineer

Ma'ed Technologies

Founded a campus-focused startup. Launched a Flutter app that reached 2,000+ active users in its first year and built a university procurement web application that cut manual workflows by 65 percent, shaped by 250+ user research sessions.