Open to new opportunities

Odile
Abimana.

CS Graduate  ·  AI Builder  ·  Data Professional  ·  Kirkland, WA

I build things with data, AI, and code. Computer Science graduate from Seattle University with a passion for elegant solutions to real problems — from local AI systems to cloud dashboards. Rwandan roots, Pacific Northwest life.

3+ Years in Tech
3 AI Products Built
Trilingual
Odile Abimana
💡 Problem Solver
📚 Reader
⚙️ Tech Enthusiast
About Me

Engineer. Builder.
Community Person.

I'm a Computer Science graduate from Seattle University with a minor in Business Administration, born and raised in Rwanda and now based in Kirkland, WA. I'm deeply curious about how AI and data can be used to build things that genuinely help people.

Over the past few years I've worked across data engineering, software development, automation, and AI systems — from Boeing's factory floors to building local-first AI tools from scratch. I love the intersection of technical depth and real-world usefulness.

I'm also trilingual — English, Kinyarwanda, and Kinyamulenge — which shapes how I think about communication, equity, and reaching people where they are.

🤖
AI & ML Systems
Local LLMs, RAG pipelines, prompt engineering
📊
Data & Analytics
SQL, Python, dashboards, pipelines
🌍
Community
Refugee support, Congo Peace, NSBE, SWE
🎓
CS + Business
Seattle University · B.S. CS · Biz Admin minor
Life Outside Work
🎹 Piano ✈️ Exploring 🤝 Community 🫂 Friends 📚 Non-Fiction 🤖 Building AI
B.S. Computer Science
Seattle University · Class of 2024
Minor: Business Administration · Seattle, WA
🇺🇸 English 🇷🇼 Kinyarwanda 🇨🇩 Kinyamulenge
Skills
SQLPythonR TableauPower BIExcel Statistical AnalysisDashboarding Data PipelinesKPI Monitoring
OllamaRAG Systems Prompt EngineeringVector DB AWSAzure DevOps React.jsStreamlit API DevelopmentAutomation
Agile / ScrumRICE Framework RoadmappingA/B Testing KPI DefinitionStakeholder Mgmt User Research
Certifications
🏅
Certified Scrum Master
Scrum Alliance
📗
Excel Specialist
Microsoft
🐍
Python Certificate
Coding Dojo
Claude 101
Anthropic
🧠
AI Fluency: Framework & Foundations
Anthropic · UCC · HEA · Ringling College
Work

Experience & Community

Project Engineer — Contractor Tech
Caprock Global Solutions · Oklahoma City, OK
Jun 2025 – Nov 2025
Applied data-driven time-motion analysis to identify workflow trends, reducing operational redundancy by ↓10%.
Delivered KPI-focused process improvement recommendations, optimizing scheduling and delivery timelines.
Partnered with cross-functional stakeholders to streamline data collection and reporting workflows.
Software Engineer — Data & Reporting Tech
Archdiocese of Seattle · Seattle, WA
Sep 2023 – Jun 2024
Queried and analyzed datasets using SQL, enabling +30% faster business decision-making for leadership.
Built a cloud-based KPI dashboard (React.js + AWS) translating complex data into clear stakeholder-ready reports.
Cleaned and transformed data from multiple sources, ensuring accuracy across all reporting pipelines.
Defined KPIs with cross-functional stakeholders and monitored them through automated data workflows.
IT & Data Analyst Intern Tech
Boeing · Everett, WA
Jun 2023 – Aug 2023
Automated updates for 1,000+ factory workstations, improving operational efficiency by +30%.
Identified performance trends via Azure DevOps, reducing system downtime by ↓15%.
Built Python automation scripts that accelerated routine data processes by +25%.
3+Years of service
130+Event guests coordinated
Languages to reach families
Refugee Resettlement Volunteer Volunteer
World Relief & Belpres Church · Kirkland / Bellevue, WA
Ongoing
Support newly arrived refugee families with housing resources, education enrollment, job search, and community integration — bridging gaps in formal services through direct, ongoing relationships.
Leverage trilingual capability (English, Kinyarwanda, Kinyamulenge) to reach families that English-only services cannot, removing barriers at one of the highest-need moments of people's lives.
Annual Event Planner Volunteer
Congo Peace Non-Profit Organization · Bellevue, WA
2023 – Present
Lead end-to-end planning of the annual "Spirit of Congo" fundraising event (130+ guests) — managing vendors, logistics, timelines, and on-site coordination under pressure with limited resources.
Affiliations
⛪ Bellevue Presbyterian Church ⚙️ NSBE 👩‍💻 SWE 🌐 AnitaB.org 🎤 Grace Hopper Conference
Projects

Things I've Built

AI tools, data platforms, and product explorations — built to solve real problems.

🤖 AI Product

Luxa AI — Local AI Marketing Platform

End-to-end, locally powered AI platform that generates luxury ecommerce video ads, voiceovers, and short-form social content for Shopify products. Built privacy-first with zero cloud API cost.

100%Local inference
6-stepAI pipeline
$0API cost
PythonOllamaStreamlit Prompt EngineeringTTS
GitHub →
Luxa AI ✨
Create viral AI video ads
Product Name
stanley water bottle
Target Audience
gym lovers
Generate Full Video Ad

Features

  • AI script generation via local LLM
  • Text-to-speech voiceover pipeline
  • Automated video with transitions
  • Shopify product image workflow
  • TikTok & Instagram output

Architecture

  • Ollama — local CPU inference
  • Zero cloud dependency
  • Streamlit UI for non-technical users
  • Python automation backend
  • Modular 6-step pipeline

Why it matters

  • Eliminates monthly API costs
  • Full video ad in minutes
  • 100% private — data stays local
  • Built for Shopify micro-brands
🧠 AI Product

Lumina AI — Document Intelligence Assistant

Privacy-first AI document assistant — upload a PDF, get intelligent summaries, and ask natural language questions answered directly from the document via a full 9-step RAG pipeline running entirely offline.

9-stepRAG pipeline
LocalLLM inference
100%Private
PythonOllamaRAG EmbeddingsVector DBStreamlit
GitHub →
Lumina 📚
Your AI study assistant
Upload a PDF
What did Odile study in college?
PDF loaded ✓
Answer: B.S. Computer Science at Seattle University with a minor in Business Administration.

Pipeline

  • PDF Upload → Text Extraction
  • Text chunking (overlapping)
  • Embedding generation
  • Semantic search on question
  • Local LLM → Answer display

Architecture

  • Ollama local LLM inference
  • Vector DB for retrieval
  • Sentence embedding pipeline
  • Streamlit interactive frontend
  • No cloud dependency

Use Cases

  • Document Q&A for research
  • PDF summarization for students
  • Privacy-compliant enterprise use
  • Legal / HR document analysis
📋 PM Case Study

YouTube Discovery Optimization

A product management case study on how YouTube's algorithm buries subscribed content. Designed "Smart Subscription Feed v2" with RICE-scored prioritization, 4 user personas, A/B test roadmap, and projected impact metrics.

+25%Subscription CTR
+30%Creator reach
162RICE score
RICEUser Segmentation A/B TestingRoadmapping
YouTube Content Discovery
IS BROKEN
2B+
Monthly Users
70%
Algorithm Views
+25%
Sub CTR
12wk
Roadmap

Problem

  • 70% views from algorithm, not subscriptions
  • Creators see 50–90% subscription view drops
  • Users feel stuck in engagement loops
  • Niche content buried by mainstream

Solution

  • Smart Subscription Feed v2
  • Priority Inbox for top subscriptions
  • Feed Filters: Subs Only, No Shorts
  • Engagement Weighting by watch history
  • Transparency: "Why am I seeing this?"

Methodology

  • 4 user segments identified
  • RICE score: 162 (P0 priority)
  • 3 solutions with tradeoff analysis
  • 12-week delivery roadmap
  • North Star: Subscription CTR +25%
📊 Data Engineering

Vitality Dashboard — Archdiocese of Seattle

Cloud-based KPI dashboard built for the Archdiocese of Seattle's bi-annual data collection, translating complex multi-source data into clear stakeholder reports — enabling leadership to make faster, better-informed decisions.

+30%Faster decisions
AWSCloud hosted
React.jsAWSSQL Data PipelinesKPI Design
📈
Vitality Dashboard
React.js + AWS + SQL cloud reporting platform
Contact

Let's build
something together.

Based in Kirkland, WA — open to new opportunities, collaborations, and conversations. I typically respond within 24 hours.

Send a message

Drop me a note — I'd love to connect.

✓ Sent! I'll be in touch soon.