Ontita Nyamusi Moraa
Nairobi, Kenya
Jomo Kenyatta University of Agriculture and Technology
Email: nyamusiontita@gmail.com
Python
JavaScript (React)
OpenAI API
Hugging Face Transformers
GitHub
Jupyter Notebooks
Matplotlib
Langchain
Docker
Kubernetes
Jupyter Notebooks
VS Code
Pinecone
AWS
TensorFlow
PyTorch
Scikit-learn
Keras
React.js
Flask
Celery
Continuous Integration/Continuous Deployment (CI/CD)
Unit Testing (PyTest, Unittest)
Code Review Practices (pull requests, merge requests)
Reporting and Documentation
Design Thinking
Remote collaboration tools
About me (50 words)
Generative AI Engineer with a strong foundation in Machine Learning (ML) and Natural Language Processing (NLP). Proficient in Python, JavaScript (React), SQL, and prompt engineering. Skilled in utilizing OpenAI API, Hugging Face, and vector databases to develop advanced AI solutions. Experienced in fine-tuning Large Language Models (LLMs) and leveraging multiple GPUs for high-performance NLP tasks, including those in Amharic, Swahili, and Yoruba languages. Expertise in deploying complex AI-driven applications using Docker and orchestrating parallel tasks with Celery. Proven ability to design and implement sophisticated generative AI models, enhancing creativity and automation in diverse projects.
Education
10 Academy ( May 2022- August 2022 )
Worked on real world challenges in the domains of Data Engineering, ML and Generative AI Engineering
Notable projects include developing a Contract Q&A Retrieval-Augmented Generation (RAG) system, creating a Precision RAG Pipeline, fine-tuning a BERT model with Amharic data, applying causal inference for logistic optimization, constructing a data warehouse for traffic data storage, and conducting sentiment analysis.
Linear Algebra
Mathematics & Algorithms
Calculus
Statistics
OOP (Object Oriented Programming)
Projects
Designed and implemented a Retrieval-Augmented Generation (RAG) system for Contract Q&A, enhancing legal professionals' efficiency and accuracy. This innovative solution integrates LLMs with external data, utilizing advanced chunking and indexing techniques to provide precise, context-rich answers to complex contractual queries, revolutionizing contract
Developed Precision RAG to enhance enterprise-grade LLM applications. The solution focuses on Automatic Prompt Generation, Evaluation Data Generation, and Prompt Testing and Ranking, streamlining prompt engineering, boosting LLM efficiency, and improving accuracy for various business needs.
Worked in a group of 4 to fine-tune an LLM that enhances customer support for African languages. The project involves creating AI systems for Amharic focusing on quality text embedding and generation to deliver relevant and efficient responses, leveraging advanced LLM models fine-tuned with extensive language datasets.
Developed a Redash chatbot add-on to enhance data analytics and visualization capabilities. This LLM-based tool enables natural language queries to extract insights from Redash dashboards and connected databases, automating SQL query generation and dashboard creation. It streamlines data exploration and empowers users with non-technical backgrounds to access and analyze complex data efficiently.