Davina Titus
Software Engineer  ·  Portfolio
Generative AI R&D · Ernst & Young
Davina Titus

Software engineer building AI-powered full-stack systems with React, TypeScript, Python, Node.js, and LLM evaluation workflows. EY Generative AI R&D intern. Builder of ParallelYou, ChiMetrics, and data-driven tools for real users.

View My Work Get in Touch Download CV
Davina Titus
University
UIC- B.S. Computer Science Software Engineering Concentration
GPA & Honours
3.54 · Dean's List, College of Engineering
Expected Graduation
May 2028
Scroll
About Me

Building with precision,
designing with purpose

"The best software is invisible — it simply works, beautifully and reliably."

I am a Computer Science student at the University of Illinois Chicago specializing in software engineering, AI systems, and human-centered technology. At Ernst & Young, I worked on Generative AI by structuring evaluation datasets, comparing GPT-4, Claude, and Llama outputs, and refining prompt workflows across enterprise automation use cases.

My strongest work sits at the intersection of full-stack engineering, data visualization, and LLM-powered products. I like building systems that are technically solid, visually polished, and easy for people to actually use.

Strongest Stack
React · TypeScript · Python · Node.js
Focus
AI Systems · Data Viz
Education
UIC CS · Software Engineering
Recognition
Dean's List · College of Engineering
3
Deployed Apps
15+
GenAI Use Cases
500+
Structured Data Items
2028
Expected Graduation
✦
Chicago, Illinois
Open to SWE and AI Opportunities
⌁
Professional Experience

Where I've Contributed

Summer Intern
Ernst & Young
June – August 2025
Bangalore
  • Built and organized 500+ structured evaluation data items for a GenAI automation platform spanning 15+ enterprise use cases.
  • Benchmarked GPT-4, Claude, and Llama outputs to identify failure patterns, compare response quality, and document 50+ targeted prompt and configuration improvements.
  • Collaborated with senior engineers to refine system behavior, prompt structure, and evaluation workflows for more consistent AI responses.
OpenAI GPT-4ClaudeLlama Prompt EngineeringLLM EvaluationGenAIDataset Architecture
Selected Projects

Crafted Work

Data & Systems
✦
ChiMetrics | Chicago Civic Data Dashboard

A live civic data dashboard for Chicago, visualizing real-time public datasets across crime trends, CTA transit ridership, red light and speed camera violations, and 311 service requests — all sourced from the Chicago Open Data Portal. Grew out of an earlier C++ red light camera analysis tool, rebuilt as a full React application with interactive charts and live API integration.

ReactViteRechartsSocrata APIData Visualization
Architecture
React/Vite frontend consuming Chicago Open Data APIs with reusable chart components and dataset-specific transforms.
Engineering Challenge
Normalized messy civic datasets into clean visual states for crime, transit, camera, and 311 trends.
Live Demo ↗ View on GitHub →
→
Artificial Intelligence
II
ParallelYou | Multiverse Simulator

An AI-powered app that generates 3 distinct life trajectory timelines based on your background, goals, and risk tolerance — powered by the Claude API. Features salary progression charts, side-by-side profile comparisons, PNG export, and dark/light mode. Fully deployed and live.

ReactViteClaude API AnthropicData VisualizationVercel
Architecture
Claude API workflow generates structured timeline data that feeds comparison cards, charts, and exportable visual outputs.
Engineering Challenge
Designed prompts and UI state to keep three generated life paths distinct, comparable, and readable.
Watch Demo ↗ View on GitHub →
→
III
FaceCart | Skincare Platform

Built at SparkHacks 2026 - a selective, competitive hackathon with a team of four. A full-stack e-commerce platform with AI-powered skincare analysis, engineered entirely in 24 hours. A custom conflict detection engine analyzes 40+ products across 200+ ingredient combinations with 0–100 compatibility scoring. GPT-4 multi-modal analysis reduced user research time by 80%.

ReactTypeScriptNode.js ExpressOpenAI GPT-4Tailwind CSS
Architecture
Full-stack React/Node app with product data, AI analysis, and compatibility scoring across ingredient combinations.
Engineering Challenge
Built a conflict-detection layer that compares skincare ingredients and turns analysis into a simple 0–100 user-facing score.
View on GitHub →
→
↗
More on GitHub

An expanding portfolio of academic projects, personal experiments, and engineering explorations.

github.com/davina-titus
→
Technical Skills

Tools & Technologies

Languages
PythonC++C JavaScriptTypeScript SQLHTML/CSS
Frameworks & Libraries
ReactNode.jsExpress Tailwind CSSPandasNumPy
AI & Data
OpenAI GPT-4ClaudeLlama Prompt EngineeringLLM Evaluation MySQLData Visualization
Methods & Tools
GitAgileUI/UX Design REST APIsOOPResponsive Design
Strongest Technical Identity
AI full-stack engineering · LLM evaluation · civic data visualization · human-centered systems
Beyond the Code

A Little More About Davina

Davina's artwork ✦
Art & Entrepreneurship

During COVID, I launched an art print business and donated a portion of every sale to ACCESS.

♩
Keyboard

Started keyboard at 8. By 12, I was performing at a professional level and earned Trinity College London Grade 3 certification.

◎
Track & Badminton

Years of track and badminton taught me how to stay composed under pressure, adapt quickly, and work toward a shared goal.

✈
Travel

I’ve explored France, Netherlands, Singapore, Australia, New Zealand, Bali,Thailand and many more - experiences that sharpened how I observe people, spaces, and culture.

◈
Fashion & Gaming

I like blending chic and casual with bold prints - a personal style that reflects how I think about contrast, detail, and expression.I enjoy fast-paced strategy games like Clash Royale and Brawl Stars, which keep me sharp, competitive, and adaptable. Reached 12k trophies on Clash Royale.

Get In Touch
Connect on LinkedIn GitHub
LinkedIndavina-titus GitHubdavina-titus Email