Languages
- TypeScript
- Python
- JavaScript
- SQL
- C++
- Java
- Solidity
Backend Engineer.AI Systems. Distributed Systems.
I'M NATHANAEL CHERAMLAK WOLDE. A software engineer, open-source contributor, and competitive programmer. I specialize in taking on critical, real-world challenges and engineering elegant solutions that truly matter.
Skills
Technologies and engineering strengths reflected across production-style AI, backend, and full-stack projects.
Backend & APIs
- Node.js
- Express.js
- FastAPI
- Flask
- Django
- REST APIs
- WebSockets
AI / ML Systems
- RAG Pipelines
- LlamaIndex
- Qdrant
- Transformers
- Computer Vision
- XGBoost
- Symbolic AI
Data & Infrastructure
- PostgreSQL
- MongoDB
- Redis
- Neo4j
- Prisma
- Docker
- Celery
Frontend & Product
- Next.js
- React
- Tailwind CSS
- Vite
- NextAuth
- OAuth
- Cloudinary
Systems & Security
- JWT Authentication
- Role-Based Access Control
- Rate Limiting
- Background Jobs
- AES-256-GCM
- Hardhat
- Ethers.js
Experience
Roles and milestones from my engineering journey building production software.
2025 - Now
AI Software Engineer
iCog Labs
November 2025 - Present | Addis Ababa
Building AI systems in production for cognitive architectures and vision-based emotion understanding.
- Developed and optimized components of OpenPsi, improving motivation and emotion models for autonomous decision-making in OpenCog Hyperon AGI.
- Deployed a production emotion-detection service with Vision Transformer and transfer learning, reaching a 92.5% F1 score with sub-200ms inference latency.
2025
AI Software Engineer Intern
iCog Labs
June 2025 - October 2025 | Addis Ababa
Focused on improving model quality and operational performance for NLP and computer vision services.
- Optimized NLP and computer vision models using advanced Transformer and CNN architectures, improving classification accuracy by 10%.
- Architected and deployed the model API with Docker and LRU caching for reliable, low-latency service portability.
2025
Backend Engineer Intern
Kuraz Tech
April 2025 - August 2025 | Remote
Improved backend reliability and speed through database, authentication, and caching optimizations.
- Designed and optimized database schemas, improving query performance by 30%.
- Implemented authentication and caching mechanisms, reducing server response time by 20%.
Education
Academic training, backend specialization, and high-volume problem-solving practice.
BSc. Computer Science
Addis Ababa University
Ranked among the top universities in Africa.
- Data Structures and Algorithms
- Object-Oriented Programming
- Operating Systems
- Intro to Software Engineering
- Computer Architecture
- Database Design
- Web Development
Competitive Programming
Africa to Silicon Valley (A2SV)
A2SV has helped 80+ students secure internships and roles at companies like Google, Palantir, and Bloomberg.
- Solved 1,000+ DSA problems across LeetCode and Codeforces.
- Achieved a Top 20 National Ranking through rigorous practice.
Software Engineering (Backend Specialization)
ALX
- Python
- Web Development
- APIs and Backend Development
- Web Servers and Security
Projects
Selected projects focused on secure architecture, real-time systems, and production implementation quality.
Showing 6 of 20
IngestIQ
RAG Document Intelligence Platform
Built a production-style document intelligence platform that ingests PDFs, indexes them into a vector knowledge base, and delivers grounded answers through a modular Retrieval-Augmented Generation workflow.
- Designed a full ingestion and query pipeline with FastAPI, Inngest, Gemini embeddings, Qdrant, and LlamaIndex to keep document processing, retrieval, and generation clearly separated.
- Added both Streamlit and React interfaces so the system supports operational ingestion workflows as well as end-user document Q&A in a reliable, product-ready experience.
- FastAPI
- Python
- Inngest
- Qdrant
- Gemini API
- React
Project Links
Mission4Nation
Digital Ministry Backend Platform
Built a modular backend platform for a church-centered digital ecosystem, supporting content delivery, subscriptions, digital commerce, donations, events, and community engagement through a production-style TypeScript architecture.
- Designed multi-domain backend workflows for sermons, books, subscriptions, testimonies, donations, live sessions, and admin reporting with role-based access and entitlement checks.
- Integrated Firebase, Cloudinary, Chapa, and cron-based lifecycle jobs to support identity, notifications, media storage, payment verification, and reliable background automation.
- TypeScript
- Node.js
- Express.js
- MongoDB
- Firebase
- Cloudinary
Private build. Demo and source available on request.
Leaderboard System
High-Performance Ranking API
Built a production-minded leaderboard API that combines PostgreSQL for durable score history with Redis for low-latency ranking, delivering fast and reliable standings for competitive applications.
- Designed a dual-storage approach where PostgreSQL preserves immutable score events while Redis sorted sets power global and per-game leaderboards across daily, weekly, monthly, and all-time windows.
- Implemented JWT authentication, Lua-scripted ranking updates, rate limiting, and automated tests to support trustworthy ranking behavior under real backend conditions.
- Node.js
- Express.js
- PostgreSQL
- Redis
- JWT
- Vitest
Project Links
Smart Scheduler
Symbolic Task Planning System
Built a full-stack task scheduling platform that models dependencies as symbolic knowledge and generates execution-ready schedules through a hybrid React, Flask, and MeTTa architecture.
- Designed a dependency-aware planning workflow with cycle detection, graph visualization, layered schedule generation, and frontend decision support for actionable task sequencing.
- Integrated MeTTa as a declarative reasoning layer so scheduling logic is expressed over symbolic task relationships rather than hidden entirely inside imperative backend code.
- React
- TypeScript
- Flask
- Python
- MeTTa
- Symbolic AI
Project Links
Hotel Management System
Full-Stack Hospitality Platform
Built a multi-role hotel management platform that brings reservations, room operations, guest services, and AI-assisted support into one reliable full-stack system designed around real hospitality workflows.
- Designed secure guest and admin experiences with JWT authentication, booking conflict checks, room availability management, and relational data modeling through Prisma.
- Integrated a Gemini-powered assistant with function-calling workflows to handle hotel tasks such as room lookup, booking support, and service requests within the product experience.
- React
- TypeScript
- Express.js
- Prisma
- SQLite
- Gemini API
Project Links
Movie Recommendation System
Graph-Based Recommendation Engine
Built a graph-powered recommendation engine with Python and Neo4j that turns user behavior, genres, tags, and movie relationships into explainable recommendations backed by a credible end-to-end data pipeline.
- Designed multiple recommendation strategies, including collaborative and content-aware approaches, to improve relevance across different user and discovery scenarios.
- Modeled recommendation logic in Neo4j with Cypher and surfaced path-based explanations, making the output more transparent, debuggable, and trustworthy.
- Python
- Neo4j
- Cypher
- pandas
- MovieLens
Project Links
Awards & Achievements
Recognitions that reflect competitive excellence, continuous learning, and execution.
Contact Me
Contact Details
- Phone: +251900017309
- Email: nathanaelcheramlak7@gmail.com