Spencer Nyaenya

SPENCER NYAENYA AI Engineer | Software Engineer | Data Analytics --https://github.com/spencers20 Nairobi, Kenya SUMMARY Dynamic and versatile AI Engineer and Full Stack Developer with expertise in AI engineering, full-stack software development, and data analytics. Proven ability to design, build, deploy, and monitor AI models, software systems, microservices, and data pipelines at scale. Skilled in leveraging Large Language Models (LLMs), vector embeddings, and Retrieval-Augmented Generation (RAG) techniques to build intelligent search and conversational AI solutions. Proficient in Next.js, Node.js, React, Flask, and experienced with containerization and orchestration using Docker, Docker Compose, and Kubernetes. Proficient in data analysis and visualization with python, using Microsoft Fabric and Power BI. Recognied for a proactive mindset, strong problem-solving abilities, and the ability to work effectively within cross-functional and collaborative team environments SKILLS • Programming Languages: Python, JavaScript, TypeScript, HTML, CSS • AI & Machine Learning: OpenAI, LangChain, Langgraph, Langsmith, Embeddings, Vector databases, Prompt Engineering, RetrievalAugmented Generation (RAG) ,LLM finetuning • Web Development: Next.js, Node.js, React, Flask, Fast,RESTful APIs, HTML, CSS • Containerization & Orchestration: Docker, Docker Compose, Kubernetes • Cloud & Deployment: Vercel, Microsoft Azure, CI/CD (GitHub Actions) • Data Engineering & Analytics: Microsoft Fabric, Power BI, Data Pipelines, Python • DevOps & Monitoring: LangSmith, Azure OpenAI Tracing, Model Observability, Logging, API Management • Databases: Postgresql, Mongodb, Redis • Other Tools: Git, Postman, VS Code, Make, n8n,zapier EXPERIENCE Data Analyst and Machine Learning Ujao Technologies Nairobi, Kenya 02/2024 - 11/2024 • Analyzed and visualized user interaction data from the g-Tahidi AI chatbot using Power BI, generating actionable insights on user engagement trends and frequency of use. • Designed and implemented a log-parsing algorithm to identify and categorize user types, enabling the marketing team to focus targeted advertising efforts, which contributed to a 5% increase in chatbot usage. • Developed and integrated AI-powered virtual assistants using Flowise into company websites, significantly improving customer support responsiveness and overall user experience. • Conducted prompt engineering and fine-tuning of AI models to enhance the quality and relevance of chatbot responses, leading to higher user satisfaction and increased interaction rates. PROJECTS Lexifile bot Lexifile Personal Project A Next.js web-based AI document analysis platform that allows users to upload documents like PDF, Word, txt, CSV, and interact with their content via natural language chat. Integrated Azure AI language models with LangChain orchestration and Pinecone vector storage for fast semantic search. Implemented document parsing, embedding generation using Cohere, and conversational querying to deliver real-time, context-aware responses. Deployed the full-stack solution on Vercel for scalable and low-latency performance. AI-Agent (Vector Query Microservice) Lexiagent Personal Project Developed a Dockerized microservice AI agent using LangChain, FastAPI, and Python, designed to accept Pinecone credentials (API key, environment, index, and namespace) via API and intelligently retrieve and answer user questions from the specified vector store. In addition to document analysis, the agent performs structured data analysis and generates visualizations based on user queries. The system leverages LangGraph for multi-step reasoning, LangSmith for observability, and is deployed on Azure for scalable and efficient access. NLP Text Classification https://colab.research.google.com/drive/1caVHUT7Jo-5CjfY-2Hm_4viWo3zQXn03?usp=sharing Personal Project Developed and evaluated an NLP text classification pipeline using Python in Google Colab. Performed data cleaning, preprocessing, and exploratory analysis on text datasets. Trained and compared multiple classification models, including Multinomial Naive Bayes (MNB) and Complement Naive Bayes (CNB), using key performance metrics like accuracy, precision, recall, and F1-score. Conducted model testing on new, unseen data to assess generalization and classification performance. Gained hands-on experience in end-to-end NLP workflows, model selection, and evaluation techniques. AfyaSphere – Health Management Platform Collaborative A web-based health application that empowers users to manage their health profiles, receive hospital reports, and interact with an AI-powered symptom checker. The system leverages a Groq AI model for fast and intelligent health query processing. Integrated Docker for containerized deployment Postgres and MongoDB for scalable data storage. Technologies used include Next js , React, Node.js, Flask, Fast, Postgre and MongoDB. EDUCATION Bachelor of Computer Science Nairobi, Kenya 09/2021 - 04/2025 (awaiting Kenyatta University graduation) Kiambu, Kenya K.C.S.E 01/2017 - 04/2021 Alliance High School Kisii, Kenya K.C.P.E 01/2009 - 11/2016 Precious Hope School INTERESTS Cloud Engineering AI and Deep Learning Motivated by a deep interest in building scalable, secure, and reliable cloud-based systems. Actively exploring cloud infrastructure, DevOps practices, and deployment strategies to deliver high-performance applications in production environments. Passionate about leveraging Artificial Intelligence and Deep Learning techniques to build intelligent systems that learn from complex data and deliver real-world impact. Strong interest in developing AI-powered solutions that enhance automation, decision-making, and user experiences across various industries. Keen on exploring advancements in NLP, LLMs, computer vision, and AI-driven system design to solve challenging business and technical problems.