Tarikua Misganaw
AI Engineer
I’m an AI Engineer with 2 years of experience at iCog-Labs, working on graph intelligence and reinforcement learning. I improved Beam Search performance, increasing pattern discovery accuracy and reducing failed reconstructions. I contributed to AI-powered graph motif interpretation systems.I led development of a certificate-driven RL framework for safe skill reuse.I have also built self-verifying AI Agentic Research Assistant with grounded retrieval and explainable AI pipelines. I’m pursuing ML/AI Engineer roles focused on scalable intelligent systems and reliable AI infrastructure.
Skills
Projects
Built a production-ready self-verifying RAG application using FastAPI, FAISS, Gemini API, and React to generate grounded answers from PDF documents with confidence scoring, citation tracking, and dynamic clarification routing for hallucination-resistant AI responses.
View Source Code
Built a production-grade RAG document assistant using FastAPI, FAISS, SentenceTransformers, and React to process PDF documents and deliver context-aware answers with adaptive retrieval, page-level citations, and sub-millisecond vector similarity search.
View Source Code
Built an AI-powered motif interpretation system for a graph intelligence platform using Gemini API, MeTTa, NetworkX, and React to generate topology-aware, hallucination-resistant explanations for mined graph motifs and enable interactive exploration of knowledge graph patterns.
View Source CodeLed the development of a certificate-driven reinforcement learning framework using PyTorch and MO-Gymnasium to generate, validate, and store compositional skills through CDS/PDS certification tests, enabling auditable and negative-transfer-resistant skill reuse for multi-objective RL environments.
View Source Code
Developed a RAG-based chatbot using Mixtral-8x7B and LangChain to deliver context-grounded health advice.
View Source Code
Built a real-time NLP web app using RoBERTa to classify tweet sentiment with confidence scoring.
View Source CodeExperience
icog-labs
12/2024 – present
Contributed to NeuroGraph AI Assistant using Graph Neural Networks and search algorithms. Designed a Gemini interpreter with topology-first prompting and built a Dockerized orchestrator for a production REST API. Enhanced Beam Search for pattern discovery and contributed to the Economic Attention Allocation system using C++ and MeTTa.I lead the development of SubRep, a certificate-driven reinforcement learning framework for auditable and negative-transfer-resistant skill reuse using multi-objective RL.
Trip ways
06/2023 – 09/2023
Developed an e-commerce website using React, Firebase, and Supabase. Implemented secure authentication methods and integrated a secure payment gateway through PayPal.
Afronex Tech hub
04/2024
Developed a responsive MERN-stack blog website with features like user authentication and dynamic post creation. Ensured seamless cross-device experience and secure session management.
Eskalate
05/2024 – 08/2024
Built a scalable blog platform backend using Go and MongoDB. Implemented CRUD operations, advanced search, and role-based access control, and integrated AI-powered content suggestions.
Certifications
Machine Learning Specialization
03/10/2024
Completed a comprehensive machine learning specialization covering supervised learning, unsupervised learning, neural networks, recommender systems, and reinforcement learning. Gained hands-on experience building and evaluating ML models using industry best practices for solving real-world AI problems.
Software Engineering Program
Completed a Google-backed 1,000+ hour software engineering program focused on Data Structures and Algorithms, algorithms, and backend development given by Africa to Silicon Valley academy.
Neural Networks and Deep Learning
25/11/2024
Studied the foundations of deep learning, including neural network architecture, forward/backpropagation, activation functions, and gradient descent optimization. Built and trained deep neural networks while developing a strong understanding of how modern AI systems learn and generalize from data.
Education
Addis Ababa Science and Technology University
BSc in software engineering • 2019– 2024
3.59/4.0
Achievements & Awards
Judge & Mentor – HackIndia 2025
2025Evaluated AI/ Web3 student projects at Chitkara & Lloyd Institute, guiding teams toward socially impactful solutions.
Semi Finalist – A2SV Internal Hackathon
Created a platform to directly connect farmers with merchants, reducing intermediaries and inflation in agricultural goods.
Get in Touch
Interested in collaboration or just want to say hello? Feel free to reach out!