I design and build end-to-end systems that connect cloud infrastructure, AI models, and automation workflows into real products that solve real problems.

What started as a pursuit of a Junior Cloud Engineer role quickly evolved into something bigger: I was entrusted to build and deploy AI-powered systems from scratch — from backend APIs and orchestration layers to vector databases and LLM integrations.

I thrive at the intersection of DevOps, backend engineering, and applied AI — where voice turns into insight, workflows become autonomous, and data becomes intelligence.

A production-grade, AI-powered system that captures raw voice notes, processes them through a semantic pipeline, and enables meaning-based search and retrieval.

• 🎙️ Capture: Record and upload audio directly from the browser.
• 🧠 Process: Transcribe audio, enrich it with a Large Language Model (Gemini), and structure the content intelligently.
• 🔎 Search: Store embeddings in Qdrant for semantic search — retrieving information based on meaning, not just keywords.

• 🌐 Frontend: Simple browser interface built with HTML, CSS, JavaScript, leveraging navigator.mediaDevices.getUserMedia for voice capture.
• 🐍 Backend: Python Flask API handling requests, routing, and serving results.
• 🧠 Processing Layer: memory_manager.py converts text to embeddings with sentence-transformers and communicates with Qdrant for vector storage.
• ⚙️ Orchestration & AI: Integrated with n8n for event-driven workflows and Gemini LLM for structuring and enrichment.

📦 Repo: Voice Memory Assistant
🎥 Demo: Watch on YouTube

• DigitalOcean, Ubuntu Server, Docker, Docker Compose, Linux Administration
• System Architecture, Workflow Automation, API Orchestration (n8n)
• CI/CD Fundamentals, Syncthing File Synchronization

• LLM Integration (Gemini via OpenRouter, Whisper)
• Semantic Search & Text Embedding (sentence-transformers)
• Vector Database Management (Qdrant)
• Context Engineering, Prompt Design

• Python, Flask, RESTful API Design
• JSON Payload Handling, Routing, Data Processing

• HTML5, CSS3, JavaScript (Vanilla)
• Media Capture APIs (getUserMedia), Fetch API (async requests)

• Problem Solving, System Design, Debugging, Independent Learning
• Using LLMs as Development Copilots for debugging, context generation, and architectural decision-making

I'm actively seeking ** opportunities** as an AI Systems Engineer or Cloud & Automation Engineer, where I can bring this mix of skills to help teams build and scale intelligent systems.

If you're working on LLM-powered platforms, automation pipelines, or semantic search tools, let's connect.

📫 Reach out:

• LinkedIn: https://www.linkedin.com/in/tafara-rugara-0627b819b/
• GitHub: https://github.com/98Devops
• YouTube: https://www.youtube.com/@techwithtaf