Grace Raper

AI Engineer / Full-Stack Software Engineer based in Seattle, WA
Website / Email / GitHub / LinkedIn

👩‍💻 Engineering Experience

Founder / Lead Engineer @ Remme (2025 - Current)
Generative AI-powered reminiscence therapy platform providing emotionally intelligent, photo-guided conversations for individuals with memory challenges.

• Reminiscence Therapy Agent: Designed and prototyped an agentic system that analyzes user photo databases to select meaningful and positive images based on past experience. Built automated session planning that generates 20-minute therapy outlines with curated photos and contextual questions, delivered through integrated video AI agents (Tavus).
• Product Development & User Research: Conducted extensive customer discovery through events at senior centers and memory care facilities. Collaborated with professors and medical professionals at UW, USC, and UMass to validate therapeutic approach and explore research partnerships and grant funding opportunities.

Founder / Lead Engineer @ SortMyShoebox (2024 - Current)
SortMyShoebox is an application that organizes and predicts metadata for digitized family photos.

• Computer Vision Pipeline: Built comprehensive image analysis system using CV libraries for face detection, age estimation, image segmentation, and labeling (TensorFlow, PyTorch, OpenCV). Integrated with MLLM models to embed image attributes for semantic searches (Hugging Faces, OpenAI embeddings).
• Agentic Photo Organization & Caption Generation: Created agentic workflows for detecting related images from events by analyzing visual similarity, faces, clothing matches, and temporal patterns. Integrated LLMs with visual analysis to generate contextual captions. Developed prompt engineering techniques combining visual elements with family history for personalized narratives.
• AWS Cloud Architecture: Designed production infrastructure using EC2, S3, Lambda, and MongoDB. Implemented identity and security frameworks (IAM), authentication protocols (SAML, OAuth, OpenID Connect), and encrypted sensitive user data (e.g., images).
• MLOps & Deployment: Built CI/CD pipelines for ML model deployment using Docker, Hugging Face, Replicate, and GitHub Actions.

Lead Software Engineer / COO @ Ghost Pacer (2021 - 2023)
Ghost Pacer is AR-Fitness startup that produced custom augmented reality glasses for outdoor applications. While I was there, we grew to seven figures in annual revenue & received funding from Techstars before being bought by Peloton in 2024.

• Advanced Algorithm Development: Designed algorithms that accurately update the user’s position every frame by combining data from embedded GPS, IMU, accelerometer, NASA elevation surveys, and previous run history. Developed ML algorithms for indoor functionality without GPS.
• Lightweight Augmented Reality Application: Implemented and optimized the app for our embedded platform with comprehensive unit testing and regular profiling analysis.
• Bluetooth Integration: Integrated Bluetooth classic and BLE peripheral devices and developed custom data protocols to communicate between the headset and iOS/watchOS applications, leveraging insights from CoreLocation and HealthKit (e.g., user heart rate, sleep statistics, walking cadence). Implemented audio functionality to connect Bluetooth headsets to receive audio workouts and verbal coaching.

Software Development Intern @ Unify Square (June 2019 — September 2019)

• Collaborated with CTO on real-time analytics dashboard prototype with REST API integrations across enterprise platforms (Zoom, Teams, Slack)
• Developed data analysis solutions using Python, R, and SQL informing strategic decisions.
• Prototype expanded into company product offering, contributing to $152M acquisition in 2021.

👩‍🎓 Education

B.S. in Computer Science @ The University of Washington (September 2019 - March 2022)

Concentration in Data Science w/ minor in Applied Mathematics.

👩‍🔧 Projects

Caspir Camera - check it out on TestFlight!

iOS photography app using Swift and Core ML for pose estimation and composition guidance to help users replicate professional photography techniques.

RedViz - https://grace-raper.github.io/redviz/

Interactive data visualization framework built with D3.js exploring correlations between historical redlining and modern census data for social justice research.

Portfolio Website - https://graceraper.com

Interactive showcase featuring data visualizations and technical projects with responsive design and performance optimization

Technical Skills

Languages: Python, Go, Java, JavaScript, TypeScript, SQL, R, Swift
AI / ML & LLMs: PyTorch, TensorFlow, scikit-learn, OpenCV, LangChain, LlamaIndex, Hugging Face, OpenAI APIs, embeddings, semantic search, RAG, prompt engineering
AI Developer Tools: Claude Code, Windsurf, Lovable
Cloud & MLOps: AWS (SageMaker, EC2, S3, RDS, Lambda, DynamoDB), GCP, Azure, Docker, CI/CD, GitHub Actions
Backend & Frontend: Django, Flask, React, REST APIs
Databases: PostgreSQL, MongoDB, Neo4j, graph databases