About
I came from Uganda to the United States with questions and resolve. In May 2026 I earned a B.S. in Data Science from the University of Mary Washington, a first-generation degree built on family sacrifice, community, and steady work.
Employers know me for shipping AI that teams can trust: schema-valid pipelines, LLM evaluation and guardrails, PII/PHI redaction, and clear stakeholder communication. Mantra: one day at a time.
The Journey
I started my internship search as an international student on an F-1 visa, where extra documentation requirements and employer hesitation made an already competitive process harder. Early applications were rough. I didn’t yet know how to present my skills, pitch myself, or use LinkedIn as more than a listing board.
The turning point was becoming a Student Alumni Ambassador in spring 2023. That role put me in rooms with alumni and mentors who helped me rebuild my resume around the right experience, sharpen my elevator pitch, and learn the most underrated skill in a job search: following up with genuine, lasting follow-through. UMW’s Center for Career and Professional Development backed that growth with workshops and one-on-one advising.
That discipline paid off directly: a spring 2025 internship with Banduri and a summer 2025 Business Intelligence internship with Navy Federal Credit Union, both sourced through networks I had to learn how to build. Now, as a Rise Peer Mentor and Resident Assistant, I pay it forward, helping other students, especially fellow international and first-generation students, navigate the same path.
Success is rarely a solo journey. There are people who genuinely want to see you thrive.
Experience
Administered a corporate OpenAI GPT training environment for Department of War senior leaders and corporate executives, providing hands-on help desk support and enterprise licensing management. Built CI/CD pipelines in Microsoft Azure DevOps and grew hands-on with JavaScript, HTML, and Vue on the frontend.
Built AI-powered compliance automation for regulated data workflows. Python/FastAPI ingestion with MinIO quarantine storage, PostgreSQL metadata, SHA-256 hashing, RabbitMQ publishing, and idempotent artifacts. Designed job state machines, Redis worker contracts, schema validation, and async pipelines. LLM enrichment via Ollama and AWS Bedrock with structured JSON, PII/PHI redaction, and executive PDF/HTML compliance exports.
Analyzed credit and debit card transactional data to understand how Active Duty members’ spending changes during Permanent Change of Station (PCS) moves, 2019–2024. Built Databricks pipelines with PySpark, developed Power BI dashboards, and delivered stakeholder presentations with actionable member-support recommendations.
Contributed to Jade Rabbit, an AI-driven healthcare product, in banduri’s inaugural internship cohort with UMW College of Business. Applied Change Architecture and agile design thinking to surface pain points, validate problem–solution fit, and deliver roadmap insights to executives.
Concurrent leadership across residence life, alumni relations, mentorship, and orientation. Active in African Student Union, NAACP, National Society of Collegiate Scholars, COAR volunteering, and Pi Mu Epsilon.
Education & Credentials
B.S. Data Science
African Student Union · NAACP · National Society of Collegiate Scholars · Resident Assistant · Student Alumni Ambassador · Orientation Leader · COAR volunteer · Peer mentor · Pi Mu Epsilon
Projects & Research
2nd Place · UMW Eagle Egg Pitch Competition
AI-detection tools guess whether writing "looks like" AI after the fact, an adversarial arms race that punishes honest students and is trivial to spoof.
Built proof-of-process instead of detection: timestamped writing checkpoints and revision history sealed into tamper-evident PDFs, with field-level encryption, Argon2 password hashing, and signed JWT sessions.
2nd place at the UMW Eagle Egg Pitch Competition; shipped as a live product on FastAPI, Next.js, and PostgreSQL.
Presented at NCUR 2026 · Richmond, VA
Emergency alerting regulatory documents are dense and scattered across agencies, making it slow for staff to find grounded, citable answers during time-sensitive decisions.
Built a retrieval-augmented generation pipeline over the regulatory corpus, Pinecone vector embeddings for retrieval, grounded generation with source citations, exposed through both a Streamlit UI and a CLI.
Selected to present at the National Conference on Undergraduate Research (NCUR) 2026 in Richmond, VA.
Local businesses want AI automation but have no low-risk way to see it working on their own use case before committing budget to it.
Shipped five production-style agentic assistants, restaurant, real estate, med spa, home services, and law, each with Claude tool-calling, an Ollama fallback, local embeddings via ChromaDB, an MCP server, and an operator admin dashboard. Guardrails cover 23 prompt-injection patterns plus session and rate limiting.
Five distinct verticals live simultaneously on Railway, the core sales tool for the AI Alo consulting practice.
Skills
Click chips to pin highlights · 0 selected
AI Alo Demos
Functional AI automation for local businesses. Claude with Ollama fallback, local embeddings, open source. Try Aria live below, no separate tab required, then explore all five.
Live
Restaurant · Aria books reservations and answers menu questions 24/7
Live
Real estate · Scout qualifies leads and powers a mini-CRM dashboard
Live
Med spa · Luna recommends treatments and screens candidacy
Live
HVAC & trades · Rex dispatches jobs and quotes from live data
Live
Law firm · Vera handles intake, SOL triage, and booking
Beyond the resume
Select a tile to learn more, same curiosity that fuels the engineering work.
Explore a tile
Contact
Open to AI/ML roles, innovation teams, and thoughtful collaborations. I respond the same day.
