// Hello, my name is
I'm a software engineer and technologist focused on building scalable cloud native applications, intelligent and production ready software designed for long-term impact. My work spans backend engineering, cloud infrastructure and AI integration across technology-driven markets.
Dynamic Software Engineer with expertise in backend and cloud engineering. Proven ability to deliver scalable software solutions and optimize system performance.
Experienced in building high-throughput production applications serving millions of daily requests. Excels in agile environments, contributing to project success through effective communication and teamwork.
Pursuing a Master of Science in Information Technology at Carnegie Mellon University Africa, with a B.Sc. in Computer Science from Karatina University.
$ whoami
paul_mutemi
$ cat profile.json
{
"role": "Cloud & Software Engineer",
"location": "Nairobi, Kenya",
"focus": ["APIs", "Microservices", "AI/LLM"],
"cloud": ["AWS", "GCP", "Azure"],
"available": true
}
$ _
Across production systems serving millions of users, I integrate large language models, computer vision, and speech AI into real business workflows — not prototypes. Every integration below is live in a shipped product.
Integrated OpenAI GPT-4, Claude, and Google Gemini APIs into production iOS backends handling grammar correction, AI chat, and intelligent document processing for 532,000+ active users.
Built Retrieval-Augmented Generation systems using ChromaDB, vector embeddings, and sentence-transformers to ground LLM responses in large-scale knowledge bases — deployed at New Age serving millions of daily requests.
Orchestrated AI pipelines with OpenAI Whisper for large-scale audio transcription and sentiment analysis, processing thousands of concurrent audio tasks per minute. Also integrated neural TTS and Wav2Lip for lip-synced avatar generation.
Built an OCR pipeline using Azure Computer Vision and OpenCV for real-time image-based currency recognition within a live App Store application, enabling offline and online image processing at scale.
Used LangChain to orchestrate multi-step AI pipelines — from audio ingestion through transcription, sentiment scoring, and structured output — replacing manual data processing workflows and improving throughput by 45%.
Designed and built fully local agentic desktop assistants using Ollama, RAG, and streaming LLM inference — zero API-key dependency, on-device inference with document-grounded responses for privacy-first enterprise use cases.
A productivity app elevating English grammar — sentence corrector, plagiarism checker, spell check, document editor, and punctuation checker. Backend integrates OpenAI, Claude, and Gemini models for intelligent text and document processing.
Live exchange rate calculator with offline mode for travelers. Built an OCR pipeline using Azure Computer Vision for image-based currency recognition and integrated real-time currency APIs for accurate conversions worldwide.
Desktop AI assistant (PyQt5) that answers from a PDF knowledge base via RAG (ChromaDB + local embeddings), Ollama for LLM inference, Whisper for speech-to-text, neural TTS, and Wav2Lip for lip-synced avatar — fully on-device, zero API keys.
End-to-end cloud-native platform built for CMU's Cloud Computing course. Engineered a Spark/Scala ETL pipeline processing large-scale Twitter datasets into structured MySQL schemas, a high-performance Go web service containerised with Docker, and full infrastructure-as-code provisioning via Terraform and Kubernetes — deployed across auto-scaling clusters.
Complete e-signature workflow and sales tax returns application for US residents. Handles individual, S-Corp, C-Corp, partnerships, and fiduciary returns across multiple states with Google Drive and Microsoft 365 integration.
A comprehensive offline English dictionary covering 500,000+ words with meanings, synonyms, antonyms, pronunciation audio, and daily word features — accessible anytime without an internet connection. Built for students and professionals looking to expand vocabulary and improve writing.
Real-time voice translation across 100+ languages — speak naturally and get instant translations without typing. Designed for international travelers, business professionals, and language learners. Features conversation history, offline support, and a clean interface that eliminates language barriers mid-conversation.
An AI-powered writing assistant integrating ChatGPT-4 for students, writers, and professionals. Generates structured paragraphs, essays, emails, stories, and poems. Includes grammar and style checking, mathematical equation solving, and AI image generation — all in one productivity platform.
Open to Backend and Cloud Engineering roles. Whether you have a project, an opportunity, or just want to connect — my inbox is always open.