Kaleab Gulilat’s autographKaleab Gulilat
Open to work
Addis Ababa, Ethiopia

Full-Stack & AI Engineer building production systems.

I design and ship products end to end — from application architecture and backend systems to polished interfaces, real-time experiences, and AI agents.

Recently, I’ve been building a production multi-tenant AI classroom combining live voice, multi-agent orchestration, RAG, assessments, and real-time student interaction. I use Mastra for more structured, observable, and maintainable agent systems.

TypeScript · Next.js · Node.js · PostgreSQL · Mastra · WebRTC · LLMs / RAG

01 / Selected work

Selected engineering work

Production products and technical systems I’ve helped architect, build, and ship across AI, real-time applications, SaaS, commerce, and business software.

AI Super Teacher — Real-time AI classroom platform interface
Real-time AI classroom platform

AI Super Teacher

Sole Full-Stack Engineer

Architected and built a production, multi-tenant classroom platform in which AI agents conduct live lessons and interact with students in real time.

The platform combines multi-agent orchestration, WebRTC voice interaction, RAG over educational materials, dynamic assessments, classroom engagement tools, personalized student insights, and organization-level authorization.

  • Next.js
  • Node.js
  • PostgreSQL
  • Mastra
  • WebRTC
  • LLMs
  • RAG
View case study
EasyBoatDeals — Marketplace & commerce platform interface
Marketplace & commerce platform

EasyBoatDeals

Full-Stack Engineer

Worked across the stack on a Payload CMS and Next.js marketplace for buying and selling boats.

Improved server-side rendering and application performance, optimized data access, implemented dynamic search and authentication flows, and integrated Stripe payments with webhook handling.

  • Next.js
  • Payload CMS
  • Stripe
  • SSR
View case study
HeyMilo AI — AI-powered interview platform interface
AI-powered interview platform

HeyMilo AI

Frontend Engineer

Contributed to the interview experience and interviewer-management application of an AI recruiting platform.

Implemented reusable Next.js interfaces, worked through a major UI overhaul, and integrated VapiAI into the candidate interview experience.

  • Next.js
  • TypeScript
  • Tailwind CSS
  • VapiAI
View case study
Nettowelt — Insurance operations platform interface
Insurance operations platform

Nettowelt

Lead Frontend Engineer

Led frontend implementation and contributed to the architecture of a PostgreSQL-backed insurance-management application built with Ruby on Rails.

Worked across real-time updates, document generation, external API integrations, role-based authorization, and complex business workflows.

  • Ruby on Rails
  • PostgreSQL
  • Hotwire
  • Stimulus
  • ERB
News Flow — High-volume content ingestion pipeline interface
High-volume content ingestion pipeline

News Flow

Full-Stack Engineer

Built an automated content pipeline processing 5,000+ articles per day from multiple sources.

Designed Puppeteer scraping and scheduled ingestion supporting as many as 15 concurrent sources, followed by processing, categorization, and AI-assisted summarization and aggregation.

  • Node.js
  • Puppeteer
  • CRON
  • AI processing
View case study
Comparobot — AI-assisted document comparison interface
AI-assisted document comparison

Comparobot

Primary Frontend & Backend Engineer

Built the primary frontend and backend of a Rails prototype where users upload energy-plan documents and receive AI-assisted comparisons and recommendations.

Worked alongside an AI engineer on agent integration and provisioned production infrastructure using Kamal on a Hetzner VPS.

  • Ruby on Rails
  • AI agents
  • Kamal
  • Hetzner
Explore the complete project archive
02 / Experience

Ownership, end to end.

AI Super Teacher

Sole Full-Stack Engineer

Own the architecture and development of a production multi-tenant AI classroom. Use Mastra for agent orchestration, with specialized teacher and supervisor agents, tools, and workflows. Designed the agent architecture, real-time classroom interactions, WebRTC voice communication, RAG workflows, background assessment generation, analytics, and organization-level authorization. Responsible for frontend, backend, data architecture, AI integration, infrastructure decisions, and production delivery.

Independent / Remote

Full-Stack Engineer

Nov 2024–Present

Work with product teams on production applications across AI, SaaS, commerce, and CMS-driven products. Recent work includes EasyBoatDeals with Next.js, Payload CMS, search, authentication, SSR, and Stripe, and HeyMilo’s AI interview product with Next.js, Tailwind CSS, and VapiAI.

Rago Software Solutions PLC

Full-Stack Engineer

Jun 2023–Aug 2024

Delivered web applications and CMS-driven products using Next.js, Payload CMS, Craft CMS, Tailwind CSS, and Ruby on Rails. Led frontend delivery while contributing to backend architecture, real-time systems, API integrations, document generation, AI prototypes, deployment infrastructure, and high-volume content ingestion.

DAF TECH Social

Full-Stack Engineer

Jun 2022–Oct 2022

Built an issue-tracking application using C# and React and contributed across planning, technical solution design, implementation, and testing for client projects.

03 / Capabilities

Full-stack product engineering

I work across the entire product stack rather than limiting myself to a single layer.

Core
TypeScript · JavaScript · React · Next.js · Node.js · PostgreSQL
AI & Agent Systems
Mastra · LLM integrations · Agent orchestration · Multi-agent architectures · RAG · Tool calling · Agent workflows · Structured generation · Context engineering · AI observability and evaluation
Backend & Realtime
NestJS · Express · REST APIs · WebSockets · WebRTC · Ruby on Rails · Payload CMS
Data & Infrastructure
PostgreSQL · MongoDB · MySQL · SQLite · Docker · Git · Kamal · AWS · Hetzner · Netlify
Additional experience
Astro · React Native · Craft CMS · Python · C# · Tailwind CSS
04 / About

Curiosity across the stack.

I’m a full-stack and AI engineer based in Addis Ababa, Ethiopia.

I enjoy working on products where frontend, backend, infrastructure, AI, and product decisions intersect rather than treating them as isolated disciplines.

My experience spans AI products, real-time systems, SaaS platforms, commerce, CMS-driven applications, internal business tools, data pipelines, and early-stage prototypes.

More recently, I’ve become particularly interested in the architecture behind production AI agents: how agents use tools, how deterministic workflows and model reasoning should interact, how context and retrieval are managed, and how complex agent behavior can be observed and evaluated.

I use Mastra as a core part of the architecture for new agentic products and the continued evolution of my existing AI systems.

Outside software, I spend a lot of time playing chess, reading, learning, and training.

Education

B.Sc. Electrical and Computer Engineering

Addis Ababa Science and Technology University · 2023

05 / Contact

Have something interesting to build?

I’m open to full-stack and AI engineering opportunities, particularly around agentic AI, SaaS, real-time systems, and technically challenging web products.

If you’re building something where I could help with architecture, implementation, or taking a product from idea to production, get in touch.